framework,version,device,op_name,kernel_source,gemm_dtype,m,n,k,latency
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,65536,12288,7.6472261216905375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,65536,16384,10.264502631293402
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,65536,10240,6.677962832980686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,12288,10.259385002983942
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,16384,14.367477416992188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,10240,8.06863742404514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,65536,8192,6.0434693230523004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,8192,6.121692233615452
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,10240,2.278425852457682
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,8192,1.8642230563693578
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,16384,3.8091634114583335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,12288,2.756636513604058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,7168,1.6168222427368164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,8192,6.136749267578125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,7168,5.242006937662761
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,65536,7168,4.45799297756619
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,10240,8.404902988009983
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,6144,1.4489644368489583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,6144,4.031681908501519
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,65536,6144,3.7172533671061196
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,12288,11.855032179090712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,5120,1.1511661741468642
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,65536,5120,3.4025501675075955
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,6144,4.353087107340495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,4096,0.9219111336602105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,5120,3.6615503099229603
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,5120,3.5979966057671438
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,65536,4096,2.4689244164360895
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,4096,2.9025743272569446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,3584,0.7845662434895834
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,3584,2.527351167466905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,65536,3584,2.8906976911756725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,65536,3072,1.926947487725152
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,3584,2.595228407118056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,3072,0.7559066878424751
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,3072,2.1601617601182723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,16384,18.956685384114582
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,65536,2560,1.6235404544406469
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,3072,2.287271923489041
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,2560,0.6043279965718588
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,2560,1.8056354522705078
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,2560,1.893413331773546
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,2048,0.5493724611070421
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,2048,1.4226293563842773
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,2048,1.5760444005330403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,1536,0.40865598784552676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,1536,1.1304595735337999
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,65536,1024,0.6973919868469238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,65536,2048,1.308808856540256
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,65536,1536,1.0503013398912218
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,1024,0.39278311199612087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,1024,0.8801866637335883
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,1024,0.8639778031243218
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,65536,768,0.5558551152547201
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,768,0.2731040053897434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,65536,512,0.3926897843678792
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,768,0.6426284578111436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,65536,256,0.2893599934048123
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,512,0.2736168967352973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,768,0.905210706922743
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,512,0.5587920082939996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,512,0.6952364179823133
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,65536,128,0.22045511669582793
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,256,0.22273688846164277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,256,0.4801644219292535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,65536,64,0.1984106699625651
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,65536,128,0.21329066488477919
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,128,0.5021831194559733
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,65536,32,0.1946106619305081
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,64,0.458238230811225
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,65536,32,0.4766213099161784
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,16384,65536,10.240208943684896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,65536,5.540167066786025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,16384,16384,2.728271060519748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,16384,5.416740417480469
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,65536,24.682403564453125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,16384,1.1913386450873482
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,16384,12288,2.110892401801215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,16384,4.382603539360894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,128,0.3180470996432834
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,12288,3.410930633544922
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,12288,0.8844488991631402
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,12288,3.1348550584581165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,16384,10240,1.88702392578125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,10240,0.6782488822937012
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,7168,5.021263122558594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,10240,2.4988702138264975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,10240,2.6155367957221136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,8192,0.4895191192626953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,16384,8192,1.6257955763075087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,1536,1.1415884229871962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,8192,1.7916143205430772
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,8192,2.0124231974283853
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,16384,7168,1.4115724563598633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,7168,0.4140044318305121
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,256,0.4154328770107693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,16384,6144,1.0858710606892903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,7168,1.5465884738498266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,7168,1.7077884674072266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,6144,0.3589048915439182
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,6144,1.3285253312852647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,6144,1.420744048224555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,16384,5120,0.9351796044243706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,5120,0.30637865596347386
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,5120,1.0767733256022136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,4096,0.2432488865322537
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,5120,1.186279084947374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,65536,4096,2.8454390631781687
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,4096,1.0088435278998482
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,16384,3584,0.6335822211371528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,4096,0.9675093756781684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,3584,0.21702399518754747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,3584,0.761531564924452
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,3072,0.18823556105295816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,16384,3072,0.6096577644348145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,3584,0.8635857899983724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,3072,0.649762683444553
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,16384,2560,0.39124356375800234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,3072,0.7441057629055448
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,16384,4096,0.6036817762586806
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,2560,0.16555910640292698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,16384,2048,0.31638844807942706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,2048,0.16171022256215414
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,2560,0.5695644484625922
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,16384,1536,0.24297777811686197
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,1536,0.11291466818915473
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,2048,0.5029555426703559
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,2560,0.6433377795749241
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,16384,1024,0.17033510737948945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,1536,0.3653457694583469
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,1536,0.427549335691664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,1024,0.086535996860928
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,16384,768,0.13691643873850504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,1024,0.25078044997321236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,1024,0.29565511809455025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,768,0.07613422473271687
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,768,0.2002168893814087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,768,0.2513057655758328
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,16384,256,0.06946933269500732
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,512,0.1974017752541436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,512,0.16066932678222656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,512,0.06833066542943318
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,16384,128,0.056306666798061795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,256,0.13789244492848715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,256,0.0634017785390218
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,128,0.12457244926028782
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,16384,64,0.047691557142469615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,128,0.14414310455322266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,16384,128,0.06316266457239787
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,2048,0.5548337830437554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,64,0.12570844756232366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,16384,32,0.13089599874284533
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,16384,512,0.103057775232527
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,256,0.15349688794877794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,16384,32,0.04895822207132975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,12288,16384,2.0923084682888455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,16384,0.881440003712972
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,65536,3.9059049818250866
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,16384,3.529735141330295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,16384,4.282806396484375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,12288,12288,1.6186933517456055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,12288,0.563410653008355
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,12288,10240,1.2820524639553494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,12288,2.819853252834744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,12288,65536,8.09248521592882
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,12288,2.6364773644341364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,10240,0.45273510615030926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,10240,2.0262834760877824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,8192,0.34972532590230304
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,12288,8192,1.0132915708753798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,10240,2.150976816813151
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,8192,1.482948409186469
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,12288,7168,0.9828355577256945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,7168,0.3161848915947808
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,8192,1.6858541700575087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,7168,1.382730695936415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,12288,6144,0.7399777836269803
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,7168,1.3936772876315648
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,6144,0.2659982310401069
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,6144,1.0886221991644967
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,12288,5120,0.6116373274061415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,6144,1.2174150678846571
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,5120,0.23288355933295357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,12288,4096,0.5275422202216254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,5120,0.896275520324707
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,5120,1.0081715053982205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,4096,0.23503022723727754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,4096,0.7224879794650607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,12288,3584,0.4052453305986192
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,4096,0.9029732810126411
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,3584,0.16773333814409044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,12288,3072,0.3883813222249349
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,3584,0.6169404453701443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,3072,0.5561546749538845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,3584,0.7323635419209799
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,12288,2560,0.28585778342352974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,3072,0.6241120232476128
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,2560,0.12654488616519505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,12288,2048,0.24529422654045951
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,2560,0.45654577679104275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,2560,0.5277511278788248
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,2048,0.39097245534261066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,2048,0.10795288615756565
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,2048,0.4493849012586806
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,12288,1536,0.175690664185418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,3072,0.17770489056905112
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,1536,0.27907199329800075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,1536,0.3421191109551324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,12288,1024,0.14492444197336832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,1024,0.20377866427103677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,1024,0.255366219414605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,12288,768,0.10199999809265137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,768,0.16062933868832058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,768,0.21096355385250518
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,12288,512,0.07256622446907891
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,768,0.0610391100247701
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,512,0.12997599442799887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,512,0.16909333070119223
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,1536,0.08566755718655056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,12288,256,0.059634665648142494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,512,0.05605600277582804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,256,0.14486666520436606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,256,0.10794933636983235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,12288,128,0.04349155558480156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,256,0.05140533381038242
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,12288,64,0.03811733259095086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,65536,14.989964803059896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,1024,0.06842488712734647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,12288,128,0.146978669696384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,128,0.0958319968647427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,64,0.09843644168641831
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,12288,32,0.05121599965625339
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,32,0.10146133104960124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,12288,128,0.05163466599252489
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,10240,16384,1.6536711586846247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,65536,3.3558684455023875
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,12288,65536,20.05492485894097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,16384,0.7451439963446723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,16384,3.750878228081597
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,16384,3.0622355143229165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,10240,12288,1.5916390948825414
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,12288,0.4480862087673611
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,10240,65536,6.7680206298828125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,10240,10240,1.2069466908772786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,12288,2.380057864718967
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,10240,0.45887019899156356
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,12288,2.346156438191732
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,10240,1.7754266526963975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,10240,8192,0.8919457329644097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,8192,0.29521510336134166
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,10240,1.9513734181722004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,8192,1.306780497233073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,10240,7168,0.8347413274976941
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,8192,1.4821555879380968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,7168,0.26092976993984646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,7168,1.1538346608479817
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,10240,6144,0.6840320163302951
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,6144,0.2303368780348036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,7168,1.2771600087483723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,6144,0.9485581715901693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,10240,5120,0.4810480011834039
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,5120,0.1957999997668796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,5120,0.796779579586453
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,10240,4096,0.44276266627841526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,4096,0.19363110595279268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,5120,0.9134630627102323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,4096,0.6282071007622613
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,10240,3584,0.3485573397742377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,16384,65536,19.126607259114582
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,6144,1.0794764624701607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,3584,0.5554018020629883
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,3584,0.6499262385898167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,10240,3072,0.30708710352579754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,3072,0.12594489256540933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,3072,0.4804417822096083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,3072,0.5693830913967556
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,2560,0.40531110763549805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,2560,0.11143022113376194
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,4096,0.7335884306165906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,2560,0.4804835319519043
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,10240,2048,0.20172266165415445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,3584,0.1473368936114841
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,2048,0.09240177604887222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,10240,1536,0.1513413323296441
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,2048,0.32613688045077854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,2048,0.396029339896308
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,10240,1024,0.1203991174697876
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,1536,0.08142399787902832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,1536,0.24782578150431314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,1536,0.3080017831590441
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,1024,0.17526311344570586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,10240,768,0.08781511253780788
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,1024,0.06049866808785332
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,1024,0.2315359910329183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,10240,512,0.06871644655863444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,10240,2560,0.23594488037957084
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,768,0.053732444842656456
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,768,0.14219644334581164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,768,0.19365955723656547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,10240,256,0.048066665728886925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,512,0.11442932817671035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,512,0.15764710638258192
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,256,0.1449928945965237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,512,0.04950222373008728
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,256,0.094687110847897
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,256,0.045350223779678345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,10240,64,0.03458933366669549
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,10240,128,0.045369777414533824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,128,0.08415555291705662
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,10240,32,0.03663377629386054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,65536,13.54292212592231
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,64,0.08470844560199314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,32,0.08683288759655422
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,10240,128,0.038125332858827375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,8192,16384,1.135823991563585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,10240,65536,16.725262112087673
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,16384,0.5056017769707574
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,65536,2.7688354916042752
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,8192,12288,0.9939439561631945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,8192,65536,5.027745988633898
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,16384,2.723175048828125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,16384,3.175124486287435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,12288,2.0627449883355036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,12288,0.353587547938029
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,10240,0.3040391074286567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,8192,10240,0.8770613140530057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,12288,2.0049707624647355
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,8192,8192,0.7100559870402018
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,10240,1.4560862647162542
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,8192,0.32362132602267796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,10240,1.7101848390367296
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,8192,7168,0.5057146814134386
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,8192,1.4262168672349718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,7168,0.21589510970645479
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,8192,1.2976897557576497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,8192,6144,0.41660888989766437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,7168,0.9989813698662652
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,6144,0.18550754917992485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,7168,1.127863989935981
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,8192,5120,0.3831813335418701
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,6144,0.8510648939344617
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,6144,0.9607867134941949
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,5120,0.1604186693827311
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,8192,4096,0.34192265404595273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,5120,0.7018542289733887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,4096,0.13037778271569148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,5120,0.8108888732062446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,4096,0.5525164604187012
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,8192,3584,0.29484711752997506
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,4096,0.6717582278781467
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,3584,0.11731199423472087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,8192,3072,0.2652960088517931
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,3584,0.5364231003655328
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,3584,0.6194257736206055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,3072,0.12447910838656956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,8192,2560,0.19331822130415174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,3072,0.5094159973992242
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,2560,0.09769866863886516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,8192,2048,0.1691733333799574
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,2560,0.35804178979661727
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,2560,0.4970844586690267
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,2048,0.3000311056772868
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,2048,0.07706133524576823
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,8192,1536,0.1324977742301093
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,2048,0.38136445151435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,8192,1024,0.09059377511342366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,1536,0.062076442771487765
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,1536,0.21901510821448433
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,1536,0.28372356626722545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,8192,768,0.0760542220539517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,1024,0.21372621589236787
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,3072,0.4230417675442166
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,1024,0.15889422098795572
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,8192,512,0.04932711190647549
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,768,0.04702311091952854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,768,0.17778666814168295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,512,0.09925599892934163
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,512,0.043269332912233144
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,8192,256,0.03639111253950331
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,65536,11.615186903211805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,256,0.08036710818608601
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,8192,128,0.030456887351142034
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,256,0.04849244488610161
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,128,0.07104621993170844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,128,0.1426888836754693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,128,0.03956977857483758
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,8192,1024,0.052184889713923134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,8192,32,0.025916443930731878
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,64,0.07076266739103529
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,768,0.12304443783230251
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,32,0.07282311386532254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,512,0.15120089054107666
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,8192,256,0.149235553211636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,8192,65536,14.436109754774305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,7168,16384,1.2676266564263237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,8192,64,0.02570488883389367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,16384,0.43864978684319395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,65536,2.2972212897406683
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,7168,12288,0.8948329289754232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,16384,2.7527741326226125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,16384,2.5046426984998917
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,12288,1.6964942084418402
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,7168,65536,4.881100548638238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,12288,0.3165760040283203
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,12288,1.9175376892089844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,7168,10240,0.8622755474514432
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,10240,0.2687457667456733
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,10240,1.318472014533149
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,7168,8192,0.7049573262532552
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,8192,0.2124577760696411
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,8192,1.0155475404527452
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,7168,7168,0.5695466465420193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,8192,1.2455724080403645
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,7168,0.21652532948387992
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,7168,0.8881111145019531
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,7168,6144,0.48302846484714085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,10240,1.5798809263441298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,7168,1.0689990785386827
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,6144,0.18239288859897188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,6144,0.8151289092169868
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,7168,5120,0.40113067626953125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,6144,0.9082418017917209
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,5120,0.14552266067928737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,7168,4096,0.3323342270321316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,5120,0.6423653496636285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,4096,0.13510933187272814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,5120,0.8240079879760742
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,7168,3584,0.25775112046135795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,4096,0.5093430942959255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,4096,0.631115542517768
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,3584,0.10827822155422634
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,7168,3072,0.22566400633917916
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,3584,0.47746753692626953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,3072,0.09325333436330159
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,3584,0.5841431087917751
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,3072,0.3911466598510742
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,7168,2560,0.18074577384524873
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,3072,0.49609867731730145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,7168,2048,0.1548373301823934
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,2560,0.08159022198783027
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,2560,0.3298604488372803
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,2560,0.41503111521402997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,7168,1536,0.1160000032848782
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,2048,0.26612888442145455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,2048,0.3527848985460069
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,7168,1024,0.08699644274181789
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,1536,0.05721244547102186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,1536,0.20078310701582167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,1536,0.2665511237250434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,7168,768,0.06737511025534736
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,1024,0.047008001142077975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,1024,0.19974311192830405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,1024,0.14206488927205405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,7168,512,0.053152889013290405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,768,0.04276266694068909
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,768,0.11382933457692464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,768,0.1666533284717136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,7168,256,0.03985688752598233
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,65536,10.58944786919488
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,512,0.08873066637251113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,512,0.03944889042112563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,512,0.1451208856370714
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,7168,128,0.029887109994888306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,256,0.07047288947635226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,256,0.03663288884692722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,128,0.062477330366770424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,7168,64,0.025447110335032146
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,128,0.036702222294277616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,64,0.06032266881730822
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,32,0.060363557603624135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,7168,65536,12.673324584960938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,256,0.14437599976857504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,7168,2048,0.06936266687181261
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,7168,32,0.026748443643252056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,6144,16384,1.1741742028130424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,65536,1.907311969333225
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,16384,2.2292658487955728
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,16384,0.41703200340270996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,6144,65536,4.109001583523221
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,6144,12288,0.7515164481268989
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,12288,0.27836444642808705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,16384,2.344890594482422
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,6144,10240,0.7057235505845812
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,12288,1.446889771355523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,10240,0.23488354682922363
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,12288,1.8139173719618056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,10240,1.1805778079562717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,6144,8192,0.5302017529805502
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,8192,0.23736445109049478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,8192,0.9385929107666016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,10240,1.4863902197943792
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,6144,7168,0.48377243677775067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,8192,1.1493333180745442
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,7168,0.20457777712080213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,6144,6144,0.37325867017110187
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,7168,0.8224497901068794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,6144,0.1491884390513102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,7168,1.0021208657158744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,6144,0.7666959762573242
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,6144,5120,0.30915734503004283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,6144,0.9007448620266385
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,5120,0.5899999936421713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,5120,0.12837689452701145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,5120,0.7258079846700033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,4096,0.47636975182427305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,4096,0.1033173335923089
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,4096,0.5990648799472386
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,6144,3584,0.21600355042351616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,3584,0.09313422441482544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,3584,0.4180764357248942
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,3584,0.5262115796407064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,6144,4096,0.24563643667432997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,6144,3072,0.20199288262261283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,3072,0.08254844612545438
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,6144,2560,0.15222044785817465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,65536,10.655467563205296
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,3072,0.3625315560234918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,3072,0.45971022711859805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,6144,2048,0.123943116929796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,65536,9.648401896158854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,2560,0.30508266554938424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,2560,0.07337511248058744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,2560,0.41722578472561306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,6144,1536,0.09901066621144612
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,2048,0.2455324331919352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,2048,0.32105422019958496
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,6144,1024,0.07150489091873169
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,1536,0.0511084430747562
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,1536,0.18476443820529512
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,1536,0.2554497718811035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,6144,768,0.05214577913284302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,1024,0.12745689021216497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,1024,0.043022221989101835
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,1024,0.18843200471666124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,6144,512,0.04351733459366692
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,768,0.15708443853590223
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,768,0.03604977660708957
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,6144,256,0.029022223419613306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,512,0.14933244387308756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,256,0.06338755289713542
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,256,0.14832977453867593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,6144,128,0.024283554818895128
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,256,0.033568888902664185
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,128,0.05578666925430298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,128,0.033752888441085815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,2048,0.06304177973005506
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,10240,128,0.14434399869706896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,6144,64,0.020724443925751578
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,6144,32,0.02145244512293074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,64,0.05345244540108574
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,32,0.053568889697392784
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,768,0.10707110828823513
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,6144,512,0.08096088965733846
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,6144,512,0.036385777923795916
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,5120,16384,0.7968426810370551
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,65536,1.573822233412001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,16384,1.825042724609375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,16384,0.3705893357594808
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,5120,12288,0.5554160012139214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,5120,65536,3.2753476036919484
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,16384,2.2049458821614585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,12288,0.23976177639431426
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,12288,1.3088578118218315
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,5120,10240,0.5506880018446181
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,10240,0.2855973243713379
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,12288,1.6556391186184352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,10240,1.0800613827175565
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,5120,8192,0.45700799094306094
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,10240,1.3665093315972223
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,8192,0.17808088991377088
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,8192,0.8568142255147299
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,5120,7168,0.37754133012559676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,8192,1.085709359910753
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,7168,0.16877955860561797
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,7168,0.772156450483534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,5120,6144,0.3204631010691325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,6144,0.14328000280592176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,7168,0.9420862197875977
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,5120,5120,0.25659110811021596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,6144,0.7366799778408475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,6144,0.8185715675354004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,5120,0.11240443918440078
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,5120,0.5441217952304417
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,5120,4096,0.21789156066046822
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,4096,0.09217244386672974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,5120,0.7280933592054578
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,4096,0.43862043486701113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,5120,3584,0.19400266806284586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,4096,0.5654488669501411
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,65536,7.9977976481119795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,3584,0.08249955707126193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,5120,3072,0.16280444463094076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,3584,0.38833067152235246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,3584,0.5021146668328179
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,3072,0.07330311006969877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,5120,2560,0.1318488915761312
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,3072,0.45873509513007266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,3072,0.3383217652638753
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,2560,0.28089600139194065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,5120,2048,0.11211733023325603
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,2560,0.4133484363555908
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,2048,0.22642132971021864
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,2048,0.05494666761822171
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,2048,0.30318845642937553
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,5120,1536,0.08311111397213407
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,1536,0.04541688826349047
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,1536,0.169286224577162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,1536,0.23847378624810112
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,5120,1024,0.058557331562042236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,1024,0.03858044412400987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,1024,0.17654311656951904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,1024,0.1200373305214776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,5120,768,0.04613777663972643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,768,0.14953333801693386
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,768,0.09631555610232884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,2560,0.06422133578194512
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,5120,512,0.03279555506176419
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,512,0.07373778025309245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,512,0.03240977724393209
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,5120,256,0.02457422183619605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,256,0.05705777804056803
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,256,0.14634399943881565
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,5120,128,0.021176000436147053
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,256,0.029659556017981634
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,65536,8.853827582465277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,128,0.04952889018588596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,128,0.029713776376512315
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,5120,32,0.020397333635224234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,5120,64,0.01971733404530419
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,64,0.04698400033844841
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,5120,32,0.046524445215861
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,5120,768,0.03558133376969232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,512,0.15174221992492676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,4096,16384,0.6205520100063747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,65536,1.2314684126112196
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,4096,65536,2.7271253797743054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,16384,0.26027199957105845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,4096,12288,0.5334151056077745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,16384,1.9545084635416667
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,12288,1.1761822170681424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,12288,0.20133511225382486
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,16384,1.5671990712483723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,12288,1.4639920128716364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,4096,10240,0.416470209757487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,10240,0.21666399637858072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,10240,0.9680560429890951
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,10240,1.229721811082628
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,65536,6.975944095187717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,8192,0.14053689108954534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,8192,0.7770995563930936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,4096,7168,0.2917626698811849
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,4096,8192,0.3782613277435303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,8192,0.9932515886094836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,7168,0.12424355083041722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,7168,0.6832666926913791
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,4096,6144,0.26643021901448566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,65536,7.734476725260417
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,6144,0.11069244146347046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,4096,5120,0.2023928827709622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,7168,0.9268009397718641
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,6144,0.5899599923027886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,5120,0.0973057746887207
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,4096,4096,0.17019911607106528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,6144,0.7614453103807238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,5120,0.49784088134765625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,4096,3584,0.1561759975221422
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,4096,0.07722310887442695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,5120,0.6498559845818414
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,4096,0.3988035519917806
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,4096,0.524420420328776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,3584,0.06947022014194064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,4096,3072,0.12528710895114475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,3584,0.35181866751776797
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,3584,0.46698665618896484
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,3072,0.061722669336530894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,4096,2560,0.11010844177669948
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,3072,0.3034195635053846
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,3072,0.4100373321109348
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,2560,0.05492622322506375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,4096,2048,0.0821217762099372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,2560,0.25672711266411674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,2560,0.3455546696980794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,4096,1536,0.07221600082185534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,2048,0.04729155699412028
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,2048,0.20492800076802573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,2048,0.2838711208767361
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,4096,1024,0.04978222317165799
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,1536,0.039889779355790876
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,1536,0.15284532970852321
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,1536,0.22402845488654244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,4096,768,0.037626667155159846
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,1024,0.033795555432637535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,1024,0.16494222482045492
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,4096,512,0.026661333110597398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,768,0.14573600557115343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,768,0.08708266417185466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,768,0.0306675566567315
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,4096,256,0.01924088928434584
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,512,0.06591466400358412
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,512,0.026904889278941687
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,256,0.04980444577005175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,256,0.15275643931494817
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,4096,128,0.015837333268589444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,256,0.024912888805071514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,128,0.04209422071774801
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,4096,128,0.024736889534526404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,4096,64,0.01475644442770216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,1024,0.10835555526945327
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,64,0.040045334233178034
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,4096,32,0.01514755520555708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,4096,32,0.0391226675775316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,512,0.14359733793470594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,7168,128,0.13679999775356716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3584,16384,0.650278197394477
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,65536,1.1199564403957791
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3584,65536,2.63356081644694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,16384,0.23975110054016113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3584,12288,0.5182479752434624
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,16384,1.4769964218139648
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,16384,1.8966719309488933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,12288,1.1303200191921658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,12288,0.18655021985371908
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3584,10240,0.386369784673055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,12288,1.4235466851128473
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,10240,0.15504266156090632
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,10240,0.9203031327989367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3584,8192,0.2913653320736355
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,10240,1.260823991563585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,65536,6.5017751057942705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,8192,0.1280239952935113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,8192,0.7384862369961209
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3584,7168,0.2585466702779134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,8192,0.9871644973754883
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,7168,0.12603999508751765
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3584,6144,0.21530400382147896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,7168,0.6526062223646376
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,7168,0.8546284569634331
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,6144,0.10159733560350205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,6144,0.5803155369228786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,65536,7.537089029947917
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3584,5120,0.20093244976467559
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,6144,0.764685312906901
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,5120,0.08928800291485256
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,5120,0.47089513142903644
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,4096,0.0703164471520318
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,5120,0.6295537948608398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,4096,0.3815946578979492
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3584,3584,0.1517004436916775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,4096,0.5143760045369467
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,3584,0.06405866808361478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,3584,0.3343511157565647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3584,3072,0.11378488275739883
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,3584,0.4558231035868327
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,3072,0.05617955658170912
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3584,2560,0.09995822111765544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,3072,0.2896275520324707
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3584,4096,0.14688977930280897
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,3072,0.39763111538357204
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,2560,0.050985776715808444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3584,2048,0.08407021893395318
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,2560,0.2425546646118164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3584,1536,0.06546311246024238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,2048,0.19262754917144775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,2048,0.273214234246148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,1536,0.14502488242255315
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,1536,0.21648532814449736
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,1536,0.03663644525739882
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,1024,0.16357955667707655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,1024,0.03050666716363695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3584,768,0.03660444418589274
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,768,0.08236355251736112
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,2560,0.3387742307451036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,2048,0.04353422257635328
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3584,512,0.026975999275843304
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,768,0.026706667410002813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3584,1024,0.04714133342107137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,512,0.06228799952401055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,512,0.14343289534250894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3584,256,0.01943111088540819
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,512,0.023923556009928387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,1024,0.10233866506152683
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,256,0.04563111066818237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,256,0.14590933587816027
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,256,0.0222080002228419
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,6144,128,0.14016000429789224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3584,64,0.018223111828168232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,128,0.03898933198716905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,128,0.13628444406721327
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3584,32,0.018553778529167175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,64,0.03564888901180691
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3584,32,0.03523911039034525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3584,768,0.15123021602630615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3584,128,0.016559999850061204
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3584,128,0.0218248897128635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,65536,0.9678079817030164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3072,65536,2.1138312021891275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,16384,0.22796532842848036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,16384,1.3927830590142145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3072,16384,0.49476178487141925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,16384,1.8228079477945964
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,12288,1.04768000708686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,12288,0.16884622308943006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,12288,1.3719316058688695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3072,10240,0.35868533452351886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,10240,0.8755066659715441
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3072,12288,0.3975386619567871
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,10240,0.14472088548872206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3072,8192,0.2732497851053874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,10240,1.1917493608262804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,65536,5.89231194390191
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,8192,0.11629688739776611
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,8192,0.7035217814975314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3072,7168,0.22676089074876574
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,7168,0.1025093330277337
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,8192,0.9423128763834635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,7168,0.6203617519802517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3072,6144,0.19569599628448486
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3072,5120,0.16170399718814427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,7168,0.8186506695217557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,6144,0.5313039885626899
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,6144,0.710844463772244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,5120,0.08147378100289239
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,5120,0.4692062271965875
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3072,4096,0.1394586695565118
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,5120,0.6007413334316677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3072,3584,0.125928004582723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,4096,0.06364088588290744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,4096,0.3580613401201036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,4096,0.4993733300103082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,3584,0.05803910891215006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3072,3072,0.10010133186976115
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,6144,0.0927377806769477
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,3584,0.3219590981801351
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,3584,0.4396524429321289
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3072,2560,0.08060622215270996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,3072,0.051912890540228955
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,3072,0.27349066734313965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,2560,0.04615733358595106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,3072,0.3847537835439046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3072,2048,0.0699582232369317
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,2560,0.2302640014224582
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,65536,7.1742952134874125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3072,1536,0.05261511272854275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,2048,0.18343999650743273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,2048,0.2663102149963379
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3072,1024,0.034830222527186074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,1536,0.03205777870284186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,1536,0.13549421893225774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,1024,0.09681866566340129
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,1024,0.026744888888465032
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,1024,0.15462488598293728
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3072,768,0.02641600039270189
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,768,0.14503022034962973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,768,0.07730578051673041
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3072,512,0.02159022291501363
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,512,0.05837955739763048
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,512,0.14440533849928114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3072,256,0.01585422290696038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,512,0.0221733334991667
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,2560,0.3291422261132134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,2048,0.03923199905289544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,256,0.0421324438518948
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3072,128,0.013053333593739403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,256,0.14483022689819336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,256,0.020298666424221463
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3072,64,0.012514666550689273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,128,0.03565688927968343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,3072,32,0.01293688929743237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,128,0.020145777199003432
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,128,0.13621866703033447
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,64,0.03263288736343384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,3072,32,0.03152444296412998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,3072,768,0.024051555328898962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2560,16384,0.42844533920288086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,65536,0.9641946156819662
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,3072,1536,0.20404888523949516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,16384,0.19586400190989176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2560,65536,1.6824906667073567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2560,12288,0.32818489604526097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,16384,1.3614702224731445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,12288,0.1484151151445177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,12288,1.0083110597398546
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,5120,128,0.1374853319591946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2560,10240,0.2958071231842041
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,10240,0.13145600424872503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,16384,1.7493004269070094
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,12288,1.3579627143012152
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2560,8192,0.21584622065226236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,10240,0.8256746927897135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,8192,0.10329688919915093
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2560,7168,0.17662221855587432
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,10240,1.1107688479953342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,8192,0.6661449008517796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,7168,0.09143822060690986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,8192,0.8969280454847547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2560,6144,0.17871999740600586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,7168,0.5909306738111708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2560,5120,0.13539111614227295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,6144,0.08283378018273248
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,7168,0.7937875323825412
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,6144,0.5036924680074056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,5120,0.4264471001095242
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,6144,0.687612427605523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,5120,0.07479466994603474
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2560,4096,0.10735466745164658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,4096,0.05720711416668362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,65536,5.433787451850043
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2560,3584,0.09477688868840535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,4096,0.341489765379164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,4096,0.47041776445176864
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2560,3072,0.08686577611499363
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,3584,0.3000488811069065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,3584,0.42355023490058047
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,3072,0.2589324315388997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,3072,0.04619377851486206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2560,2560,0.07519377602471246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,3072,0.3645351197984483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,5120,0.5807021988762749
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,2560,0.2175662252638075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2560,2048,0.05380533470047844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,2560,0.31473954518636066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,2048,0.03502577874395583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,2048,0.17185066805945504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,2048,0.2546479966905382
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,3584,0.05194222264819675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,1536,0.02860355708334181
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2560,1024,0.029651555750105116
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,1536,0.1270728905995687
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,1024,0.022887999812761944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,1024,0.1499048868815104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2560,768,0.022659555077552795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,768,0.06893422206242879
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,768,0.14724266529083252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2560,512,0.01721244388156467
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,768,0.021225778592957392
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,2560,0.041153778632481895
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,512,0.05338133374849955
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2560,256,0.013402666482660504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,512,0.14952266216278076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,256,0.03779022230042352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2560,1536,0.04324977927737766
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,256,0.14881155225965711
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2560,128,0.012109333442317115
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,256,0.018248000078731112
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,65536,6.84822760687934
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,1536,0.19590932793087432
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,128,0.031141334109836157
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2560,128,0.13458222813076443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2560,32,0.010593777729405297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,128,0.018216000662909616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2560,64,0.010371555884679159
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,1024,0.09077155590057373
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,64,0.028864890336990356
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2560,32,0.028320001231299505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2048,16384,0.3180684513515896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2560,512,0.020223110914230347
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,65536,0.722815990447998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,16384,0.166721781094869
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2048,12288,0.28622666994730633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,16384,1.2363750669691298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,12288,0.128438221083747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2048,10240,0.24155288272433809
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,12288,0.9286390940348307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,16384,1.6684622234768336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,10240,0.11258399486541748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2048,65536,1.3491670820448134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2048,8192,0.19001244174109566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,12288,1.2798906962076824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,10240,0.7907066875033908
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,8192,0.09045777718226115
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2048,7168,0.13664889335632324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,10240,1.0600524478488498
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,4096,128,0.14123466279771593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,8192,0.6274417771233453
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,7168,0.08084444204966228
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2048,6144,0.12965688440534803
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,8192,0.8650613360934787
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,7168,0.5511075655619303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,6144,0.07387822204165988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2048,5120,0.11523466640048557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,7168,0.7578124470180936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,5120,0.06763199965159099
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2048,4096,0.09072177939944798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,6144,0.6532124413384331
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,5120,0.3979529009924994
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2048,3584,0.06866933239830865
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,5120,0.5523564550611708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,4096,0.05094666613472832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,4096,0.32326311535305446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,3584,0.0455662210782369
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,4096,0.4572088983323839
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,3584,0.2828017870585124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,3584,0.3979955514272054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,3072,0.04010666741265191
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2048,2560,0.057838221391042076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,3072,0.24407466252644858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,3072,0.35312265819973415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2048,2048,0.04381155636575487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,2560,0.03570577833387587
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,2560,0.2045360008875529
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,65536,5.054125467936198
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,6144,0.4750080108642578
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,2560,0.297705782784356
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2048,1536,0.033435556623670794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,2048,0.16125510798560247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,2048,0.2420889006720649
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,2048,0.029693331983354356
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2048,1024,0.025104888611369666
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,1536,0.11603555414411758
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,1536,0.18723200427161324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2048,768,0.019783111082182992
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,1024,0.08300089173846774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,1024,0.020662221643659804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,768,0.06682578060362074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2048,3072,0.06609155734380086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,768,0.019243554936514962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2048,512,0.014567999376191033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,768,0.14776266945732966
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2048,256,0.011685333318180509
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,512,0.017913776967260573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,512,0.0490515563223097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,512,0.14719822671678331
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2048,128,0.010072888599501716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,256,0.034632888105180525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,256,0.017093333933088515
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,256,0.14972355630662706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2048,64,0.009734222458468543
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,128,0.029119110769695703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,128,0.1383626725938585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,128,0.016563556260532804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,2048,32,0.010015111002657149
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,64,0.02542488939232296
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,2048,32,0.025078222155570984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,65536,6.524888780381944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,2048,1536,0.025230222278171118
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,2048,1024,0.14442666371663412
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1536,16384,0.2867448859744602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,16384,0.14831911192999944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1536,12288,0.20648977491590711
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,16384,1.1626497904459636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,12288,0.11127289136250813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,12288,0.8736666573418511
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1536,10240,0.17853689193725586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,16384,1.6185564464992945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,65536,0.5896862347920736
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1536,65536,1.0020995669894748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,12288,1.2464453379313152
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1536,8192,0.1422026687198215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,10240,0.09775466389126247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1536,7168,0.12295733557807074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,10240,0.7320328818427192
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,8192,0.5877368715074327
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,7168,0.07221333185831706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,8192,0.8344728681776258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,7168,0.5161031087239584
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1536,6144,0.10221155484517415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,7168,0.7394177648756238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,6144,0.0659102201461792
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1536,5120,0.09154844284057617
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,6144,0.444485346476237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,8192,0.07787733607821994
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,6144,0.6388906902737087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,5120,0.05992888742023044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1536,4096,0.06839199860890706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,5120,0.5428266525268555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,4096,0.04503555430306328
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1536,3584,0.06104177898830838
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,4096,0.2961137824588352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,10240,1.0541662640041776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,3584,0.04047733214166429
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,3584,0.2655493418375651
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,3584,0.39202578862508136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,65536,4.690554724799262
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,3072,0.22284889221191406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1536,2560,0.04224977890650431
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,3072,0.035762667655944824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,3072,0.3445893393622504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1536,2048,0.034472889370388456
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,2560,0.031151112582948472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,2560,0.18786044915517172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,5120,0.37328534656100804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,2560,0.2920773294236925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,2048,0.15161421563890246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1536,1536,0.026335999369621277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,2048,0.23350045416090223
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1536,1024,0.01962933275434706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,1536,0.022660444180170696
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,1536,0.11096266905466716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,1536,0.18351288636525473
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1536,768,0.01590755581855774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,1024,0.07732266849941678
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,1024,0.14334132936265734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,1024,0.019084443648656208
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1536,512,0.012062222593360476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1536,3072,0.05138044555981954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,768,0.058239108986324735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,768,0.017775999175177682
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,768,0.14433511098225912
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1536,256,0.01038577738735411
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,512,0.04438933398988512
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,512,0.14304089546203613
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,512,0.016708445217874315
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1536,128,0.009296889106432596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,256,0.030701332622104224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,256,0.01551999979548984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,256,0.1452666653527154
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1536,64,0.009861333502663506
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,128,0.024737776981459722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,128,0.1367351081636217
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,128,0.015509333875444202
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1536,32,0.008605333666006723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,64,0.022657778528001573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1536,32,0.021948445174429152
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,65536,6.343578762478298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1024,16384,0.17532622814178467
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1536,4096,0.44492355982462567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1024,65536,0.698104010687934
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,65536,0.4692737791273329
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,16384,0.14549333519405788
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1536,2048,0.026888888743188646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,16384,1.086326175265842
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,12288,0.827271991305881
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,12288,0.0941937764485677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1024,10240,0.1144133276409573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,16384,1.54803646935357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1024,12288,0.13955288463168675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1024,8192,0.09357688824335735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,10240,0.0839831100569831
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,8192,0.06609777609507243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,12288,1.1709057490030925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,10240,0.6840986675686307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1024,7168,0.0843306647406684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,8192,0.5485768847995335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,7168,0.061744888623555504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1024,6144,0.07170044713550143
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,10240,0.988159073723687
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,8192,0.7978097597757975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,7168,0.4828480084737142
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,6144,0.05674755573272705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1024,5120,0.05622844563590156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,7168,0.7060666614108615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,6144,0.4158915678660075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,5120,0.052128000391854175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1024,4096,0.04842933350139194
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,6144,0.6068124241299099
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,5120,0.34908533096313477
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1024,3584,0.039992888768514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,4096,0.03848355677392747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,5120,0.5135404268900553
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,4096,0.2822933461931017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1024,3072,0.03789866632885403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,4096,0.4212471114264594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,3584,0.24811198976304796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1024,2560,0.030037333567937214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,3072,0.030380444394217596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,3584,0.37606310844421387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,3072,0.21438399950663248
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1024,2048,0.02427911096149021
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,3072,0.3289573457505968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,2560,0.026162667406929865
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,65536,4.334926181369357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,2560,0.17903110716078016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1024,1536,0.01960266629854838
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,2048,0.02225333286656274
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,2048,0.14158488644493952
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,2048,0.22632000181410047
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1024,1024,0.014319111903508505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,1536,0.10285511281755234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,1536,0.019256000717480976
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1024,768,0.011879111329714457
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,1024,0.07060977485444811
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,1024,0.016587555408477783
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,1024,0.1488328907224867
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1024,512,0.00903555585278405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,768,0.015486222174432544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,768,0.053139554129706494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,768,0.14431911044650608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1024,256,0.007270221908887227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,512,0.03854666815863715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,3584,0.034479998879962497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,512,0.14559022585550943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,512,0.014540443817774454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1024,128,0.00655377780397733
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,256,0.027638221780459087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,256,0.14832622475094265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1024,64,0.006962666908899943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,128,0.02163377735349867
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,128,0.013790222505728403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,128,0.1367502212524414
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,1024,32,0.007295110987292395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,64,0.019268444842762418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,1024,32,0.01887022289964888
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,2560,0.2789368894365099
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,768,16384,0.1395262214872572
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,1536,0.1766364441977607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,768,65536,0.5066924624972874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,1024,65536,6.043349372016059
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,65536,0.4613813294304742
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,16384,0.14053422874874538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,1024,256,0.013791110780504016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,768,16384,1.050034628974067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,768,12288,0.8114550908406576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,12288,0.09273688660727607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,768,10240,0.09767378038830227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,16384,1.5071271260579426
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,12288,1.1467662387424045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,768,12288,0.11212089326646592
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,768,8192,0.07150222195519342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,10240,0.08273155821694268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,8192,0.06554755237367418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,768,10240,0.6614586512247721
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,768,7168,0.06689511405097114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,768,8192,0.5326595306396484
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,10240,0.9631502363416883
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,768,6144,0.05624177720811632
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,7168,0.0606328911251492
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,8192,0.7770791053771973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,768,7168,0.4675600263807509
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,768,5120,0.05046844482421875
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,6144,0.055789331595102944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,7168,0.687678231133355
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,768,6144,0.40279467900594074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,768,4096,0.035819556978013777
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,5120,0.05104533169004652
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,6144,0.595653322007921
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,768,5120,0.3378569020165338
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,4096,0.037471999724706016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,768,3584,0.03167555398411221
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,5120,0.5045288933648003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,768,4096,0.27380532688564724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,3584,0.03338933322164748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,768,3584,0.240302218331231
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,3584,0.367756446202596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,3072,0.02903999884923299
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,768,3072,0.20771911409166124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,768,2560,0.02384977704948849
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,768,65536,4.162345886230469
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,3072,0.3211599985758464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,2560,0.025331555141343012
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,768,2048,0.019521777828534443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,768,2560,0.17337777879503039
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,2560,0.273144006729126
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,768,2048,0.13730311393737793
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,2048,0.021911111142900255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,2048,0.22094044420454237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,768,1024,0.012140444583363004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,768,1536,0.0994631118244595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,1536,0.018914666440751817
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,1536,0.16901332802242705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,768,768,0.010033778018421596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,768,1024,0.06911289029651217
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,1024,0.14539999432033962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,768,768,0.053039110369152494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,768,3072,0.027941332923041448
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,768,512,0.007640000018808577
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,768,0.015495111544926962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,768,0.1451751126183404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,768,256,0.006589333216349284
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,768,512,0.0366960002316369
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,512,0.014480888843536377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,512,0.14103110631306967
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,768,128,0.005882666756709416
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,768,256,0.025701334079106648
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,256,0.013801777528391944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,256,0.14361688825819227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,768,128,0.019885333047972787
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,768,64,0.00794044468137953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,128,0.13863466845618352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,768,32,0.00543200017677413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,128,0.013752000199423896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,768,1536,0.016018667154841952
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,768,64,0.01777866648303138
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,768,32,0.017100445098347135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,512,16384,0.09989866945478652
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,65536,5.900058746337891
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,512,65536,0.39797422620985246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,768,1024,0.016528889536857605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,65536,0.36931199497646755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,512,12288,0.07671378056208293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,768,4096,0.41126489639282227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,12288,0.07665599717034234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,512,10240,0.06500266657935248
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,512,16384,1.009237289428711
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,512,12288,0.760566234588623
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,10240,0.06930577754974365
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,12288,1.1131804784138997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,512,10240,0.6364320119222006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,16384,1.4714186986287434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,16384,0.09931821955574884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,512,7168,0.044771555397245616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,8192,0.05418222149213155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,10240,0.9367759492662219
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,512,8192,0.5118311246236166
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,512,6144,0.03884799944029914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,7168,0.05059555504057142
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,512,7168,0.4490959909227159
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,8192,0.7580257521735297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,512,5120,0.03168711066246033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,6144,0.04711911082267761
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,7168,0.6703839831882052
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,512,6144,0.38703110482957626
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,512,8192,0.05266666743490431
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,6144,0.5805493460761176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,512,5120,0.3251439995235867
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,5120,0.04347999890645345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,5120,0.4916568862067328
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,512,3584,0.022014222211307947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,512,4096,0.26272887653774685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,4096,0.4038906627231174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,512,3584,0.23129423459370932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,512,3072,0.01947022146648831
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,3584,0.028221332364612158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,3584,0.35931555430094403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,512,65536,4.010887145996094
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,512,3072,0.1991893317964342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,512,2560,0.01716444392999013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,512,2560,0.16666844156053331
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,2560,0.2657342222001818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,512,2048,0.014260444376203748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,2560,0.020798222886191476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,4096,0.030829333596759375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,2048,0.018162666095627677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,512,2048,0.13089777363671198
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,512,1536,0.011738667057620155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,512,1536,0.09468266699049209
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,1536,0.015684444043371413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,512,1024,0.009368889033794403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,3072,0.02389066749148899
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,3072,0.31021422810024685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,512,768,0.008032889001899296
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,1024,0.014089778065681458
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,512,1024,0.06567466921276517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,1024,0.151037335395813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,512,512,0.0069351109365622205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,512,4096,0.025310221645567153
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,512,768,0.05008444521162245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,768,0.013409777647919126
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,768,0.1475964387257894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,512,256,0.005391110976537068
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,512,512,0.034783111678229436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,512,0.14438311258951822
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,512,0.01275199982855055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,512,128,0.004855999930037392
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,512,256,0.02405599918630388
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,256,0.14522933959960938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,512,64,0.004253333227501975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,512,128,0.018263111511866253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,2048,0.21056000391642252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,512,32,0.004575110971927643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,128,0.1357626650068495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,512,64,0.01622577839427524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,512,32,0.015460444821251763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,1536,0.16684177186754015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,256,65536,0.2425831158955892
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,256,16384,0.06165066692564222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,65536,0.36161332660251194
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,512,65536,5.747737036810981
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,16384,0.10184088680479263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,256,12288,0.045090665419896446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,256,0.012065778175989786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,512,128,0.012101333174440594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,256,10240,0.038352890147103205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,256,12288,0.7397671275668674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,16384,1.4297528796725805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,10240,0.06883466906017728
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,256,10240,0.6152524418301052
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,256,8192,0.03173866536882188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,12288,1.0854596032036674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,8192,0.05361066593064202
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,256,7168,0.028227554427252874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,256,8192,0.49530580308702254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,256,16384,0.9828773074679904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,10240,0.9115236070421008
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,256,6144,0.022780444886949327
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,8192,0.7362684673733182
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,256,7168,0.434969769583808
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,6144,0.046335998508665294
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,256,5120,0.016935999194780987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,7168,0.6506089104546441
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,6144,0.5626551310221354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,5120,0.042951110336515635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,256,5120,0.31446488698323566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,256,4096,0.013985777894655863
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,5120,0.4786879751417372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,12288,0.07601955864164564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,256,65536,3.8691308763292103
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,256,3584,0.01315199997689989
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,4096,0.02962488929430644
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,256,4096,0.2543448872036404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,7168,0.05025777882999844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,256,3072,0.011395555403497485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,4096,0.39308622148301864
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,3584,0.026304889056417678
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,256,2560,0.01038044442733129
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,256,3584,0.22386222415500215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,3072,0.023792889383104112
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,3584,0.3487253453996446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,256,3072,0.19283110565609404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,256,2048,0.008993777963850234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,3072,0.3070862293243408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,2560,0.02103288968404134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,256,2560,0.16123644510904947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,256,1536,0.00759733302725686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,2560,0.2601155440012614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,256,2048,0.1262693272696601
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,2048,0.20894400278727213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,2048,0.018211555149820115
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,256,1024,0.006631111105283101
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,256,1536,0.09093333615197076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,1536,0.015994666351212394
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,1536,0.160772442817688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,256,768,0.005619555711746216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,256,1024,0.062231997648874916
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,1024,0.14749956130981445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,1024,0.014087999860445658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,256,512,0.004920888692140579
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,768,0.013434666726324292
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,768,0.1486319965786404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,256,256,0.004267555558019214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,256,512,0.033244444264305964
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,512,0.14872977468702528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,512,0.01274577776590983
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,256,128,0.003866666720973121
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,256,256,0.02260622216595544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,256,0.14432177278730604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,256,0.01202933324707879
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,256,64,0.0035208890007601846
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,256,128,0.016695999436908297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,256,32,0.003531555748648114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,65536,5.589528825547959
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,256,64,0.01442311041884952
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,256,32,0.013818666338920593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,256,6144,0.3747697671254476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,128,65536,0.18280000156826445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,128,16384,0.04689688815010918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,65536,0.3701431221432156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,256,768,0.04774577750100029
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,128,12288,0.036208000448015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,256,128,0.01202222208182017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,128,16384,0.9601564407348633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,128,12288,0.7220346662733289
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,12288,0.07767911089791192
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,128,10240,0.03003644280963474
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,16384,0.09972800148857965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,10240,0.0690444442960951
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,128,8192,0.026391999589072332
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,128,8192,0.4866959783766005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,128,10240,0.6042328940497504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,128,65536,3.795556386311849
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,8192,0.05580621957778931
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,128,7168,0.022424888279702928
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,128,7168,0.42703999413384336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,7168,0.05168088939454821
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,128,6144,0.015511110424995422
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,128,6144,0.36767644352383083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,256,128,0.1407582230038113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,6144,0.046652442879147
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,128,5120,0.011721777419249216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,128,5120,0.30886843469407826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,16384,1.4256675508287218
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,5120,0.043417778280046254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,128,4096,0.009863110880057016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,128,4096,0.24951555993821886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,10240,0.9102630615234375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,4096,0.3919457859463162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,128,3584,0.009298666483826106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,128,3584,0.2191822263929579
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,3584,0.3498942322201199
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,3584,0.027297778262032404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,128,3072,0.008266666697131263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,4096,0.030561778280470107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,128,3072,0.18891733222537568
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,12288,1.0801146825154622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,3072,0.024311999479929607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,128,2560,0.00758844448460473
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,3072,0.30407555898030597
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,2560,0.021864889396561518
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,128,2560,0.15761778089735243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,128,2048,0.006606222026877933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,7168,0.6499404377407497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,2560,0.25616444481743705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,128,1536,0.00589155571328269
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,128,2048,0.12282133102416992
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,2048,0.2033511135313246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,6144,0.5624773237440321
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,128,1536,0.08921866946750218
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,128,1024,0.004874666531880696
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,1536,0.01628533336851332
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,1536,0.16370311048295763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,128,768,0.004572444491916233
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,1024,0.014468444718254937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,1024,0.14328266514672175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,128,768,0.04520000020662943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,128,512,0.0041546668443414895
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,768,0.14283733897738987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,128,512,0.032187554571363665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,512,0.14197688632541233
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,128,256,0.003698666476541095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,512,0.01276711126168569
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,128,256,0.02197244432237413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,256,0.14188622103797063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,2048,0.018591110905011494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,256,0.012082666986518435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,128,128,0.0034906665484110513
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,8192,0.7350124253167046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,128,128,0.015847999188635085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,128,64,0.0031635556370019913
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,128,1024,0.060532444053226046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,128,0.012048888537618848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,128,32,0.0031831111345026228
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,128,64,0.013923555612564087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8192,128,768,0.013422222601042854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,128,32,0.01332266628742218
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,64,16384,0.050160888168546885
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,64,12288,0.035875555541780256
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,64,65536,0.17479022343953451
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,64,10240,0.029310223129060533
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,65536,5.559689839680989
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,64,8192,0.025035555164019268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,64,12288,0.716808001200358
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,64,10240,0.5993777910868326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,64,7168,0.021743110484547083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,64,6144,0.014172444740931192
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,64,8192,0.4821457862854004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,64,5120,0.010077333284748925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,64,4096,0.008657777474986183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,64,7168,0.4227120081583659
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,64,6144,0.36382489734225804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,64,3584,0.007988444632954067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,64,3072,0.007248000138335758
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,64,5120,0.30533332294887966
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,64,2560,0.007340444458855524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,64,4096,0.24637778600056967
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,64,3584,0.21715911229451498
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,64,2048,0.006314666734801398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,64,3072,0.18722932868533662
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,64,1536,0.005592000153329637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,64,1024,0.004570666700601578
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,64,2560,0.1564906703101264
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,64,768,0.004218666503826777
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,64,1536,0.08806488911310832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,64,1024,0.05936977598402235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,64,512,0.0038399998512532976
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,64,256,0.0034328889515664843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,64,512,0.03153866529464722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,64,128,0.003134222287270758
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,64,128,0.01515111161602868
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,64,64,0.0028515555378463534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,64,64,0.013242666920026144
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,64,32,0.002875555513633622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,64,32,0.013063111239009433
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,64,2048,0.12247199482387966
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,32,65536,0.17091555065578887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,64,16384,0.9533493253919813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,32,16384,0.04740000102255079
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,64,65536,3.769890679253472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,32,12288,0.035951998498704694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,64,256,0.020960888928837244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,32,10240,0.02793866727087233
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,32,8192,0.023788443870014612
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,32,10240,0.5992577870686849
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,32,7168,0.020584000481499564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,32,12288,0.7157422171698676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,32,16384,0.9533857769436307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,32,6144,0.012817777693271637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,32,5120,0.00965244405799442
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,64,768,0.044808000326156616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,32,4096,0.008024888734022776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,32,7168,0.42287643750508624
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,32,5120,0.30511734220716685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,32,3584,0.007248889240953658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,32,3072,0.00657066661450598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,32,4096,0.24698045518663195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,32,2560,0.006243555496136348
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,32,3584,0.2190408839119805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,32,2048,0.005277333160241445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,32,2560,0.1580559942457411
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,32,1536,0.005579555614127054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,32,8192,0.481370660993788
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,32,2048,0.12300888697306316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,32,1024,0.004560889055331548
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,32,768,0.0042168887125121225
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,32,1536,0.0882986651526557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,32,1024,0.05973866913053724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,32,512,0.003855111284388436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,32,256,0.0033048888047536216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,32,768,0.04484355449676514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,32,128,0.0031840000301599503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,32,512,0.03178933262825012
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,32,256,0.02094488839308421
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,32,64,0.002847111059559716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8192,32,32,0.0031786666562159858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,32,128,0.015171556009186639
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,32,6144,0.3643013371361627
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,32,64,0.013447999954223633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,32,32,0.012791111237472959
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,32,3072,0.18889778190188936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8192,32,65536,3.773985968695747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,16384,1.6541013717651367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,65536,12288,3.7079340616861978
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,65536,16384,4.975988599989149
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,12288,4.431194729275174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,16384,6.1400968763563375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,12288,1.174216906229655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,16384,5.896102057562934
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,12288,4.197016821967231
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,10240,1.009177737765842
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,65536,10240,3.3686169518364797
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,10240,3.5631073845757375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,8192,0.8548471132914225
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,65536,8192,3.0908417171902127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,8192,2.8799697028266062
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,10240,3.491806242201063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,7168,0.7233857578701444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,8192,2.7974906497531467
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,65536,7168,2.3548622131347656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,7168,2.4300897386338973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,6144,0.6442764600118002
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,65536,6144,2.071247100830078
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,7168,2.437925338745117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,6144,2.1074487898084855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,65536,5120,1.5756880442301433
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,5120,0.5428906546698676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,6144,2.156431198120117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,5120,1.775851567586263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,4096,0.44233155250549316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,65536,4096,1.2393982145521376
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,5120,1.7917262183295355
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,4096,1.4068951076931422
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,65536,3584,1.0962070888943143
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,4096,1.4432373046875
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,3584,0.3912399874793158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,128,0.14459555678897434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,3584,1.2296053568522136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,3072,0.34314844343397355
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,3584,1.2749857372707791
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,65536,2560,0.704528013865153
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,65536,3072,1.05733150906033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,3072,1.0526150597466362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,2560,0.2995706664191352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,3072,1.093971570332845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,65536,2048,0.6493608686659071
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8192,128,5120,0.477908452351888
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,2048,0.2530311213599311
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,2560,0.9231689241197375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,65536,1536,0.5285849041408962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,1536,0.21009333928426108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,2048,1.0597964392768013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,2048,0.7215520011054145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,65536,1024,0.3283715513017443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,1536,0.5736222267150879
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,1536,0.7298515637715658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,65536,768,0.2623866664038764
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,2560,0.8894293043348523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,1024,0.4549591276380751
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,65536,512,0.195978668000963
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,1024,0.6179848776923286
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,768,0.39153867297702366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,768,0.37485334608289933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,512,0.27770309978061253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,512,0.30641155772738987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,512,0.14347822136349148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,256,0.2382897800869412
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,256,0.2508453263176812
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,256,0.1103137797779507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,65536,64,0.09979289107852513
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,65536,128,0.18151733610365126
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,128,0.110390219423506
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,128,0.24406133757697213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,65536,32,0.0996897750430637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,1024,0.16751466857062447
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,64,0.2377999888526069
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,65536,32,0.2435377703772651
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,65536,768,0.1603040032916599
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,65536,256,0.19594311714172363
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,65536,128,0.11046044694052802
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,16384,16384,1.2341537475585938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,16384,0.4222319920857747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,16384,1.8756124708387587
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,16384,12288,0.9884729385375977
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,65536,2.0938337114122176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,12288,1.3500844107733831
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,16384,1.9295919206407335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,12288,0.3250613212585449
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,10240,0.2714471022288005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,16384,10240,0.9404097663031684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,12288,1.3847466574774847
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,10240,1.1018186145358615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,10240,1.1617493099636502
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,8192,0.2194453345404731
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,8192,0.8969466951158312
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,16384,65536,5.034284379747179
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,16384,7168,0.5715591112772623
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,8192,0.9397128423055013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,7168,0.7615448633829752
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,7168,0.20757333437601724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,16384,8192,0.6687751346164279
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,16384,6144,0.5712995529174805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,6144,0.17278577221764457
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,7168,0.8146417935689291
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,16384,5120,0.36525689231024844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,6144,0.6581137975056967
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,5120,0.1563991175757514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,6144,0.6942328876919217
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,16384,4096,0.2820275624593099
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,5120,0.7679484685262045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,4096,0.12376266055636936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,4096,0.4638231065538194
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,16384,3584,0.27777600288391113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,4096,0.4874400032891168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,3584,0.10803377628326416
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,3584,0.38605867491828066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,16384,3072,0.24486221207512748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,3584,0.4287537733713786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,3072,0.09444355302386814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,16384,2560,0.21067733234829375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,3072,0.375507566663954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,3072,0.3291155497233073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,5120,0.602135976155599
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,16384,2048,0.17684622605641684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,2560,0.29675200250413686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,2560,0.33454399638705784
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,2048,0.2237102190653483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,16384,1536,0.11901956134372288
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,2048,0.07069155242707995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,2048,0.26942043834262425
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,16384,1024,0.08952799770567153
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,1536,0.18059289455413818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,1536,0.057477335135142006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,16384,768,0.0690444442960951
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,1024,0.1323653327094184
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,1024,0.1704773373074002
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,1024,0.04875555634498596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,16384,512,0.05530222256978353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,768,0.04395022326045566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,768,0.1488648917939928
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,768,0.10582311285866632
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,16384,256,0.036068442795011736
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,512,0.08819822470347087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,512,0.04051555527581109
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,512,0.14620888233184814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,16384,128,0.03083733386463589
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,256,0.07461866405275133
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,256,0.144031114048428
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,256,0.03608622153600057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,16384,64,0.025721778472264607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,2560,0.08484711249669392
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,128,0.06968266434139676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,16384,32,0.026258667310078938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,64,0.07071733474731445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,32,0.0711324413617452
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,1536,0.22126489215426973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,65536,8.21736568874783
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,16384,65536,9.927769131130642
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,16384,128,0.0359226663907369
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,12288,16384,1.1295493443806965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,16384,0.37494399812486434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,16384,1.5438497331407335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,16384,128,0.14211822880638972
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,16384,1.6102053324381511
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,12288,12288,0.87990845574273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,12288,1.097980393303765
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,65536,1.647483507792155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,12288,0.24742044342888725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,12288,65536,3.94428719414605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,12288,10240,0.634410646226671
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,10240,0.20678933461507162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,12288,1.1666427188449435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,8192,0.168358219994439
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,12288,8192,0.4842720031738281
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,10240,0.9106800291273328
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,10240,0.9649972915649414
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,8192,0.7701830863952637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,8192,0.8206773334079318
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,7168,0.1528302166197035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,12288,6144,0.3642266591389974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,7168,0.755107561747233
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,7168,0.6897768974304199
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,6144,0.1366951068242391
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,6144,0.547393798828125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,12288,5120,0.28394844796922475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,6144,0.6099457740783691
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,5120,0.11542756027645534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,12288,4096,0.2331911192999946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,5120,0.4486515786912706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,12288,7168,0.41759289635552305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,5120,0.5311022334628636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,4096,0.09390311108695136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,12288,3584,0.20934132734934488
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,4096,0.35178399085998535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,4096,0.4162755542331272
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,3584,0.08419289191563924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,12288,3072,0.16420177618662515
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,3584,0.30547555287679035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,12288,2560,0.14751466115315756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,3584,0.3719804551866319
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,3072,0.2630106608072917
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,3072,0.3221528795030382
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,2560,0.22371111975775826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,12288,2048,0.12164089414808486
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,2560,0.06659466690487333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,2560,0.2884391148885091
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,2048,0.05630844169192844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,12288,1536,0.09415022532145183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,1536,0.14569600423177084
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,1536,0.19402221838633218
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,12288,1024,0.068031112353007
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,1536,0.046168890264299184
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,65536,7.74458991156684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,1024,0.10910311010148789
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,1024,0.14978844589657253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,12288,768,0.05180089010132683
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,3072,0.0745955573187934
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,1024,0.04436888959672716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,65536,6.713871002197266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,12288,512,0.04266400138537089
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,768,0.08705244461695354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,768,0.14832977453867593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,768,0.03571022219128079
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,12288,256,0.028911110427644517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,512,0.033152888218561806
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,512,0.14482577641805014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,12288,128,0.02390044430891673
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,512,0.08338489135106404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,256,0.1486942238277859
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,2048,0.1846026711993747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,12288,64,0.020989333589871723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,128,0.05780888928307427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,128,0.13800888591342503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,128,0.03163999981350369
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,64,0.056686222553253174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,32,0.05780888928307427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,12288,256,0.06012800004747179
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,12288,2048,0.2339279916551378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,12288,256,0.02980355421702067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,12288,32,0.021555556191338435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,16384,0.2857191032833523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,65536,1.3009618123372395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,10240,16384,0.7531404495239258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,10240,12288,0.6082631217108833
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,16384,1.3125093248155382
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,12288,0.21302578184339735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,16384,1.4601128896077473
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,10240,10240,0.4693457815382216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,12288,1.075251579284668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,10240,65536,3.2973804473876953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,12288,1.3279511133829753
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,10240,0.17900178167555067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,10240,0.8238897853427463
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,10240,8192,0.42821865611606175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,8192,0.14264621999528673
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,10240,0.9030933380126953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,10240,7168,0.32762932777404785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,8192,0.6504337522718642
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,8192,0.7285600238376193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,7168,0.1331226693259345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,10240,6144,0.2724711100260417
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,7168,0.6265048980712891
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,7168,0.6512248780992296
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,6144,0.11713866392771403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,10240,5120,0.24560711118910047
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,6144,0.4741075303819444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,5120,0.1012151108847724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,6144,0.5596515337626139
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,10240,4096,0.19495999813079834
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,5120,0.4061253335740831
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,5120,0.4720293151007758
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,4096,0.3122231165568034
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,4096,0.3801679876115587
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,3584,0.27469245592753094
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,3584,0.07250933514700995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,10240,3072,0.16017777389950222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,3584,0.33513334062364364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,3072,0.23492089907328287
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,3072,0.06511021984948052
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,10240,2560,0.12591555383470324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,3072,0.29602132903205025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,10240,2048,0.10223199923833211
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,2560,0.05553244551022848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,2560,0.19957155651516387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,2560,0.2685297860039605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,10240,1536,0.08574133449130589
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,2048,0.04821422364976671
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,2048,0.16399644480811224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,2048,0.21581333213382295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,10240,1024,0.05740711424085829
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,4096,0.08212088876300387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,1536,0.13341066572401258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,1536,0.03985777828428481
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,1536,0.17822221914927164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,65536,6.624087015787761
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,1024,0.034508443540996976
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,1024,0.14344710773891872
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,10240,512,0.033788445923063494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,768,0.14988444911109075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,768,0.031919111808141075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,512,0.06378489070468478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,65536,5.922621409098308
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,10240,256,0.024301333559883967
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,512,0.1488648917939928
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,256,0.14514933692084417
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,256,0.05609777900907728
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,10240,128,0.02438666588730282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,256,0.02664711078008016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,128,0.050387554698520236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,10240,128,0.13549066914452448
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,10240,3584,0.17080267270406088
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,10240,64,0.019211555520693462
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,10240,32,0.02014488809638553
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,10240,768,0.04790311058362325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,1024,0.09682577848434448
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,64,0.04910755488607618
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,768,0.07740800248252021
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,10240,32,0.04933866527345446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,512,0.029236445824305218
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,8192,16384,0.6163671281602647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,10240,128,0.026408889227443274
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,65536,1.1873555713229709
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,16384,1.1499706904093425
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,16384,1.2994204627143013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,8192,12288,0.7424711121453179
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,8192,65536,2.656080881754557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,12288,0.8445528878106011
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,12288,0.9527511596679688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,16384,0.25854576958550346
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,12288,0.18724000453948975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,8192,10240,0.40466756290859646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,10240,0.15389777554406062
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,8192,8192,0.36116798718770343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,8192,0.12354932890997992
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,10240,0.7085706922743055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,10240,0.8067457940843371
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,8192,0.553558243645562
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,8192,7168,0.31645332442389595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,7168,0.1126097838083903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,8192,6144,0.2444195482465956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,8192,0.697682645585802
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,7168,0.4859706560770671
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,6144,0.09862399763531154
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,7168,0.5795493125915527
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,8192,5120,0.21293066607581246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,6144,0.5292426745096842
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,8192,4096,0.15583822462293836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,6144,0.5073582331339518
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,5120,0.08389244476954143
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,5120,0.34886754883660215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,8192,3584,0.15118489000532362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,4096,0.06908533308241102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,5120,0.4480560090806749
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,4096,0.29477156533135307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,3584,0.06178577740987142
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,4096,0.37699111302693683
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,8192,3072,0.12648622194925943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,3584,0.31558577219645184
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,8192,2560,0.1025351087252299
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,3072,0.21087022622426352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,3072,0.054788443777296275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,3072,0.27010756068759495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,8192,2048,0.0815840032365587
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,2560,0.04839466677771675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,2560,0.2350133260091146
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,2560,0.1709902286529541
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,8192,1536,0.06430044439103869
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,2048,0.14371377891964382
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,2048,0.1991057793299357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,65536,5.433124542236328
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,8192,1024,0.05007288853327433
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,1536,0.11373155646853977
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,1536,0.03404533200793796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,1536,0.16343732674916586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,8192,768,0.037865777810414634
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,1024,0.14865422248840332
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,1024,0.028741333219740126
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,768,0.06888088915083143
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,768,0.14492800500657824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,768,0.026308443811204698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,65536,5.2077352735731335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,512,0.056038220723470054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,3584,0.2395999961429172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,512,0.02362488872475094
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,512,0.14470311005910239
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,8192,256,0.02196177840232849
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,8192,128,0.016418667303191293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,256,0.04478755593299866
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,256,0.021078222327762183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,256,0.14608532852596706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,8192,64,0.01441511180665758
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,8192,32,0.014900444282425774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,128,0.02090399960676829
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,64,0.04073955615361532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,32,0.039234668016433716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,8192,2048,0.04073066512743632
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,1024,0.0900684462653266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,8192,512,0.027765333652496338
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,8192,128,0.043136000633239746
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,7168,16384,0.6288257704840766
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,65536,1.0533866882324219
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,16384,0.23828532960679796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,16384,1.0539111031426323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,7168,12288,0.6670053270128039
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,7168,65536,2.548695034450955
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,12288,0.7953617837693957
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,16384,1.2412035200330946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,12288,0.1644337839550442
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,12288,0.9150213665432401
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,7168,10240,0.3482035530938043
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,10240,0.13929688930511475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,7168,8192,0.2862142191992866
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,10240,0.6523440149095323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,8192,0.11284532811906602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,8192,0.5141928990681967
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,10240,0.7839004198710123
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,7168,7168,0.2555769019656711
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,8192,0.7185395558675131
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,7168,0.1002275546391805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,7168,0.4521795378790961
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,7168,6144,0.2228142155541314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,7168,0.5567466417948405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,6144,0.0874017808172438
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,7168,5120,0.17634667290581596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,6144,0.39995021290249294
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,5120,0.3236577775743273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,5120,0.08625688817765977
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,7168,4096,0.144904891649882
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,65536,4.930013444688585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,5120,0.41166310840182835
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,4096,0.06194577614466349
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,7168,3584,0.1169368823369344
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,4096,0.25494223170810276
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,4096,0.3449386755625407
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,7168,3072,0.08803288804160224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,3584,0.2221511072582669
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,3584,0.2948382165696886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,3072,0.048954665660858154
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,7168,2560,0.08734666638904148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,65536,4.902969784206815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,2560,0.16252888573540583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,2560,0.2226053343878852
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,2560,0.04372088776694404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,7168,2048,0.060975107881757945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,6144,0.4837706883748372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,2048,0.03628622161017524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,7168,1536,0.05972088707817925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,2048,0.19147911336686876
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,2048,0.13361777199639216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,7168,1024,0.04240977764129639
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,1536,0.15508088800642225
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,1536,0.10572444730334812
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,1024,0.0790604419178433
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,3584,0.05724355247285631
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,1024,0.025623111261261836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,3072,0.19496088557773164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,768,0.0633173320028517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,3072,0.26020444764031303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,7168,512,0.026911111341582403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,768,0.023432888918452795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,512,0.050944000482559204
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,512,0.1483279996448093
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,7168,256,0.021338666478792827
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,256,0.04056266612476773
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,256,0.019398222366968792
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,7168,128,0.01719377769364251
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,1536,0.03130755490726895
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,128,0.035967999034457736
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,1024,0.14810933007134333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,128,0.14010843965742323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,7168,64,0.014128888646761576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,7168,32,0.014766222900814481
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,128,0.018956444329685636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,64,0.03532000051604377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,768,0.1438035567601522
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,7168,32,0.03533333208825853
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,7168,512,0.021010667085647583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,7168,256,0.1435662243101332
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,6144,16384,0.47453954484727645
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,7168,768,0.03426133261786567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,65536,0.8679422272576226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,16384,0.1975395547019111
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,16384,0.9690293206108941
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,6144,12288,0.4745137956407335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,6144,65536,2.030096901787652
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,12288,0.16773777537875703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,16384,1.144420411851671
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,12288,0.7196070883009168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,6144,10240,0.3352835443284776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,10240,0.12315200434790717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,12288,0.8671982023451063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,6144,8192,0.289955563015408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,10240,0.6510515742831761
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,8192,0.11279466417100693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,10240,0.7264835569593641
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,8192,0.4795991049872504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,6144,7168,0.2302844524383545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,7168,0.08834666675991482
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,6144,6144,0.19956799348195395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,7168,0.4393377833896213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,7168,0.5246026780870225
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,6144,0.0776968863275316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,6144,0.36107201046413845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,6144,5120,0.15680710474650064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,6144,0.49094401465521914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,5120,0.30262933837042916
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,5120,0.0660159985224406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,5120,0.38962311214870876
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,6144,4096,0.12241244316101074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,4096,0.05416355530420939
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,4096,0.23873866928948295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,65536,4.576669481065538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,8192,0.5944746865166558
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,4096,0.3198942343393962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,6144,3584,0.11263199647267659
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,6144,3072,0.10263911220762466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,3584,0.04901866780387031
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,3584,0.20575911468929717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,3584,0.30156177944607204
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,3072,0.17500621742672393
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,3072,0.0433555543422699
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,6144,2560,0.07776888873842028
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,3072,0.24729779031541613
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,65536,4.530430263943143
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,6144,2048,0.06313155757056342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,2560,0.03837688763936361
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,2560,0.20826755629645455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,2560,0.1567982170316908
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,6144,1536,0.05565066470040215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,2048,0.03229511115286086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,2048,0.12331467204623753
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,2048,0.18164532714419893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,6144,1024,0.036181333992216326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,1536,0.027154665854242113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,1536,0.15303644869062635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,1536,0.09744444158342148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,6144,768,0.03164711263444688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,1024,0.022691556149058874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,1024,0.14324711428748235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,6144,512,0.021765333082940843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,768,0.14638133843739828
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,768,0.02054666644997067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,6144,256,0.016255110502243042
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,512,0.04629688792758518
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,512,0.019462222854296368
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,256,0.03598755598068237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,256,0.017367111312018502
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,256,0.1425573296017117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,128,0.03221866819593642
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,128,0.1334773302078247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,6144,64,0.012672888735930124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,6144,128,0.017208000024159748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,6144,32,0.012913777596420713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,64,0.03187822302182516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,32,0.03153866529464722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,1024,0.07291111018922594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,6144,768,0.05820088916354709
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,6144,512,0.1450586716334025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,6144,128,0.013416888813177744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,5120,16384,0.44914934370252824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,65536,0.7700133323669434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,16384,0.8819715711805555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,5120,65536,1.6444809171888564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,5120,12288,0.3599875503116184
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,12288,0.1463199986351861
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,16384,1.0746648576524522
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,12288,0.6512728797064887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,5120,10240,0.2923288875155979
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,16384,0.1802355580859714
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,10240,0.12015555964575873
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,10240,0.5457715458340114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,12288,0.8319777912563748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,5120,8192,0.23896177609761557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,8192,0.08591644631491767
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,10240,0.6830915345085992
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,8192,0.46730931599934894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,8192,0.5573902130126953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,7168,0.07755377557542589
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,7168,0.3870968818664551
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,5120,6144,0.18145867188771567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,7168,0.49332443873087567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,6144,0.06713244650099012
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,6144,0.3351217905680339
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,5120,5120,0.1346044407950507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,6144,0.42450666427612305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,5120,0.058026664786868624
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,5120,4096,0.10910489161809285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,5120,0.28300266795688206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,65536,3.8885599772135415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,5120,0.3677537706163194
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,5120,3584,0.08986222081714207
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,5120,7168,0.20457243919372559
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,4096,0.21596177419026694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,4096,0.2953280078040229
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,5120,3072,0.08277777830759685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,3584,0.04220799936188591
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,3584,0.18900799751281738
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,3584,0.26351910167270237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,5120,2560,0.0680168867111206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,3072,0.03775022096104092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,3072,0.16099554962582058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,3072,0.23089867168002656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,5120,2048,0.050385776493284434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,65536,4.1647317674424915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,2560,0.033320890532599554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,2560,0.197580443488227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,5120,1536,0.04517777760823568
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,2048,0.1132871045006646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,2048,0.1677386628256904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,2048,0.028095109595192805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,5120,1024,0.02981511089536879
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,1536,0.024322667055659827
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,1024,0.06576355298360188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,1024,0.1398248937394884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,5120,768,0.021757332815064326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,1024,0.019712888532214694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,4096,0.04775822162628174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,768,0.052410665485594005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,768,0.018072888255119324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,5120,512,0.017129777206314933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,512,0.04087288843260871
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,512,0.18702488475375703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,5120,256,0.013636444177892474
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,512,0.016757332616382174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,2560,0.13605777422587076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,256,0.0333262218369378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,5120,128,0.012182222472296821
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,256,0.14221866925557455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,1536,0.1471439997355143
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,128,0.029126223590638902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,128,0.1358826690249973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,5120,64,0.010486221975750394
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,5120,32,0.011784888803958893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,64,0.02868177824550205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,32,0.028148445818159316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,5120,768,0.1483502255545722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,256,0.01573333309756385
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,5120,1536,0.08923911386066014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,5120,128,0.01577422188387977
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,65536,0.6114426718817817
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,4096,16384,0.35653244124518496
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,16384,0.17030666934119332
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,4096,12288,0.24980176819695366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,16384,0.804069307115343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,12288,0.6188622050815159
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,12288,0.107040888733334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,16384,0.9973342683580187
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,4096,65536,1.2756559583875868
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,12288,0.8851884206136068
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,10240,0.0897795557975769
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,10240,0.4988728629218207
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,4096,8192,0.18444444073571098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,10240,0.6366826693216959
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,8192,0.07208622164196439
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,4096,7168,0.16601422097947863
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,8192,0.3976711167229547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,8192,0.5163288646274143
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,7168,0.06481066677305433
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,7168,0.35579554239908856
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,4096,6144,0.1404871145884196
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,4096,10240,0.22272978888617623
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,6144,0.05928977992799547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,6144,0.30083200666639537
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,4096,5120,0.10843288898468018
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,6144,0.4019537766774495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,5120,0.05221600002712674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,4096,4096,0.08837333652708267
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,5120,0.2534044318728977
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,5120,0.33750934071011013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,65536,3.860688951280382
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,4096,0.04177688890033298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,4096,3584,0.07761689027150472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,4096,0.19736088646782768
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,4096,0.2767182191212972
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,4096,3072,0.06325688627031115
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,3584,0.036158220635520086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,3584,0.24794578552246094
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,3584,0.1713200012842814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,4096,2560,0.05860977702670627
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,3072,0.032484445306989886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,3072,0.14561511410607234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,3072,0.21156088511149088
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,65536,3.4549920823838978
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,2560,0.029309332370758057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,2560,0.18461244636111787
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,7168,0.46064445707533097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,2048,0.10123555527793036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,4096,1536,0.034338666333092585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,2048,0.15743199984232584
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,2048,0.024596444434589807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,4096,1024,0.025495999389224585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,1536,0.021553777986102637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,1536,0.14761422740088567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,1536,0.08106311162312825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,4096,768,0.01955200069480472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,1024,0.05887999799516466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,1024,0.14208088980780706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,1024,0.018030222919252183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,4096,512,0.014869333969222175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,768,0.047600001096725464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,768,0.14732889334360758
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,4096,256,0.012135110795497894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,512,0.1403084463543362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,512,0.015114666687117683
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,256,0.028935111231274072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,256,0.14080711205800375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,4096,128,0.010662222074137794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,256,0.014263111684057446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,4096,2048,0.042187554968727954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,128,0.026106667187478807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,4096,128,0.13962666193644205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,4096,64,0.009752000371615091
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,128,0.013997332917319404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,4096,32,0.010076444182131026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,64,0.025099555651346844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,32,0.025064888927671645
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,4096,768,0.01608533329433865
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,512,0.03692177931467692
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3584,16384,0.3059057659573025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,4096,2560,0.12297599845462376
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,65536,0.5269333521525065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,16384,0.13169066111246744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,16384,0.7634008725484213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3584,12288,0.24663821856180826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,12288,0.09747822417153253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,16384,0.9831920199924045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3584,10240,0.20332533783382842
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,12288,0.5611528820461696
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3584,65536,1.2354692882961698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,10240,0.0834231111738417
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,12288,0.7660977575514051
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,10240,0.46767557991875547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,8192,0.06666844420962863
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,10240,0.6475146611531576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,8192,0.37877689467536074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,8192,0.5124302440219456
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3584,7168,0.14702133337656656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,65536,3.2288292778862844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3584,6144,0.12945066557990179
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,7168,0.06080266502168444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,7168,0.32866133583916557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3584,5120,0.10317066642973159
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,6144,0.28686488999260795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,6144,0.4119253423478868
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3584,8192,0.15747643841637507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,5120,0.23875112003750273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3584,4096,0.08040799697240193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,5120,0.05090044604407417
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,5120,0.3347582287258572
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,4096,0.038345777326160006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3584,3584,0.06645066870583428
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,4096,0.1883315510219998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,4096,0.27206489774915904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3584,3072,0.060921774970160596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,3584,0.03511200017399258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,3584,0.24354399575127494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,3072,0.031110223796632554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,3072,0.21041244930691191
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3584,2560,0.05257688959439596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,7168,0.4574133555094401
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,2560,0.11772355768415664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,2560,0.02719377809100681
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,2560,0.18334844377305773
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3584,2048,0.03489777776930068
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,2048,0.02299733294381036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,2048,0.15607910686069065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3584,1536,0.029081778393851385
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,65536,3.8234206305609812
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,1536,0.07721244626575045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3584,1024,0.02260711126857334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,1536,0.019122666782803006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,3584,0.16375466187795004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,1024,0.055404444535573326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3584,768,0.01788444485929277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,1024,0.14555733733707005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,3072,0.14025689495934382
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,768,0.043679998980628125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,6144,0.05449688765737745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3584,512,0.014051554931534661
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,768,0.1460906664530436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,512,0.03438399897681342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,512,0.14886577924092612
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,512,0.014448000325096978
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,2048,0.09625777933332656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,256,0.026782222919993933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,256,0.14880443943871394
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3584,128,0.0090506664580769
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,256,0.013456000222100152
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,1536,0.1450106700261434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3584,64,0.008618666893906063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,128,0.013427555561065674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3584,32,0.009944889280531142
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,64,0.023649776975313824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,32,0.023040889037979975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,1024,0.016499555773205228
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,8192,128,0.13778221607208252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3584,768,0.015432000160217285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3584,256,0.010332444475756751
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3072,16384,0.2757999897003174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,65536,0.4798115624321832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3072,65536,0.9668604532877604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3584,128,0.02438933319515652
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3584,128,0.1393395529852973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,16384,0.12137688530815972
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3072,12288,0.20416711436377633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,16384,0.7021590868631998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,12288,0.09464889102511936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3072,10240,0.16093956099616155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,12288,0.5284524493747288
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,10240,0.07522666454315186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,16384,0.9697120454576281
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3072,8192,0.14308977127075195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,10240,0.4561955663892958
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,10240,0.6235031021965874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,8192,0.061066667238871254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,8192,0.3608720037672255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3072,7168,0.11985777484046088
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,8192,0.49156443277994794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,7168,0.056088891294267446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,7168,0.3156319989098443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3072,6144,0.10413599676556057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,7168,0.4383715523613824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,6144,0.05037600133154127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3072,5120,0.08115822076797485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,6144,0.2715724574195014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,12288,0.7190826733907064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,65536,2.964965396457248
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,5120,0.04423822296990288
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3072,4096,0.0665137767791748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,5120,0.22651200824313691
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,5120,0.32609155442979604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3072,3584,0.05589955382876926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,4096,0.034656888908810086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,4096,0.1784586641523573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,4096,0.2642310990227593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3072,3072,0.04957955413394504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,3584,0.15414044592115614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,3584,0.23579645156860352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3072,2560,0.038902223110198975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,3072,0.028237332900365193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,3072,0.13107911745707193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,3072,0.20186311668819854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3072,2048,0.03647199935383267
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,2560,0.11071377992630005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,2560,0.02514311174551646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,2560,0.1750080055660672
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3072,1536,0.027141332626342773
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,2048,0.09048266543282403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,2048,0.1514213350084093
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,65536,3.664216783311632
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3072,1024,0.019699555304315355
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,1536,0.06977688603931002
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,1536,0.14712266127268472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,1536,0.018645332919226754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3072,768,0.015569777952300178
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,1024,0.14887200461493597
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,1024,0.05163644419776069
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,6144,0.38471378220452207
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3072,512,0.01204177737236023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,768,0.0417848891682095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,768,0.14735999372270372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3072,256,0.010682666467295753
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,512,0.03219733304447598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,512,0.013746666411558786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,512,0.14630666044023302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3072,128,0.009007111191749573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,256,0.025013332565625507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,256,0.012773333324326409
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,256,0.1432524389690823
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,3584,0.031630221340391375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,3072,128,0.13812711503770617
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,128,0.012810666528013019
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3072,32,0.008605333666006723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,64,0.021943110558721755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,32,0.02162666618824005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,2048,0.02148977749877506
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,1024,0.01593866613176134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,3072,768,0.015063110325071545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2560,16384,0.226511107550727
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,65536,0.41437244415283203
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,3072,128,0.022638221581776936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,16384,0.10318844185935126
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2560,12288,0.17291822698381212
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,16384,0.6654408772786459
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,12288,0.07881333430608113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,12288,0.5121439827813042
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,16384,0.9110666910807291
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,12288,0.6924168798658582
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2560,65536,0.8747822443644205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,10240,0.42200443479749894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2560,8192,0.12455111079745823
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,10240,0.06845688819885254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,3072,64,0.010079111489984725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,10240,0.5828800201416016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,8192,0.05453688899676005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,8192,0.33897866143120664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,8192,0.47988531324598527
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,65536,2.7494746314154734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,7168,0.2992444568210178
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2560,6144,0.0907555553648207
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,7168,0.4252133369445801
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2560,10240,0.15269511275821263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2560,5120,0.07235911157396105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,6144,0.04479733440611097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,6144,0.25817155838012695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,6144,0.3660124407874213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,5120,0.03976266582806905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2560,4096,0.05472800135612488
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,5120,0.21504622035556367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,5120,0.3109955522749159
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2560,3584,0.04830311073197258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,4096,0.03133599956830343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,4096,0.16963288519117567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2560,7168,0.09777244594362046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,65536,3.498297797309028
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,3584,0.1449999941719903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2560,3072,0.04003466831313239
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,3584,0.02824977702564663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,3584,0.2280986573961046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2560,2560,0.03336800138155619
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,3072,0.025314667158656653
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,7168,0.05016266637378269
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,3072,0.19941422674391004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,3072,0.12434400452507867
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2560,2048,0.026905778381559584
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,2560,0.10329333278867935
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,2560,0.17107821835411918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2560,1536,0.022685334086418152
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,2560,0.02260977857642704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,2048,0.01940266622437371
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,2048,0.08591288990444607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,2048,0.14594844977060953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2560,1024,0.015969778100649517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,1536,0.016606221596399944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,1536,0.06596799691518147
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,1536,0.1489831076727973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2560,768,0.01276088919904497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,1024,0.014224888549910652
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,1024,0.14235022332933214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2560,512,0.01035911093155543
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,768,0.013800889253616333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,768,0.14643555217319065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2560,256,0.00795022232664956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,512,0.030496888690524634
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,512,0.012757333616415659
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,512,0.14767111672295466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2560,128,0.007275555696752336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,256,0.02332177758216858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,4096,0.25374044312371147
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,256,0.14374132951100668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2560,64,0.006605333338181178
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,128,0.020962667134073045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2560,128,0.13674844635857478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2560,32,0.00675644435816341
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,128,0.012049777640236748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,64,0.020058666666348774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,32,0.0199635558658176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2048,65536,0.5865768856472439
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,1024,0.04837422238455879
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2560,768,0.03961688942379422
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2048,16384,0.17842843797471789
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,16384,0.08929422166612412
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2560,256,0.012131555212868584
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,16384,0.6265128983391656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,12288,0.06872622172037761
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,12288,0.47347376081678605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,16384,0.8721742100185819
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2048,10240,0.12106577555338542
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,12288,0.6669395764668783
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,10240,0.06004089117050171
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2048,8192,0.09848088688320583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,10240,0.397669341829088
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,10240,0.5607288678487142
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2048,12288,0.1344204478793674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,8192,0.32023466957939994
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2048,7168,0.08685955736372207
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,8192,0.45941866768731016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,7168,0.044630222850375705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2048,6144,0.073344886302948
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,7168,0.2819848855336507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,7168,0.40916000472174746
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,6144,0.03959733247756958
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2048,5120,0.057164443863762744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,6144,0.2422835561964247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,6144,0.35602577527364093
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2048,4096,0.04619822237226698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,5120,0.03483555714289347
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,5120,0.20257333914438883
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,65536,0.3569128778245714
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,5120,0.30205244488186306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2048,3584,0.04063911239306132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,4096,0.02793866727087233
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,4096,0.15800978077782526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,4096,0.24510934617784289
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2048,3072,0.0333315564526452
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,8192,0.04855555627081129
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,3584,0.02533866630660163
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,3584,0.13754222128126356
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,3584,0.21646222803327772
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2048,2560,0.029511110650168523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,65536,3.334885279337565
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,3072,0.022808889547983806
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2048,2048,0.02460799945725335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,3072,0.18808978133731416
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,2560,0.020255110330051847
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,2560,0.1625759998957316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,65536,2.52822134229872
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2048,1536,0.018980445133315194
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,2048,0.0805964469909668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,2048,0.01772622267405192
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,2048,0.14355644914839003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2048,1024,0.014119111829333834
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,1536,0.015658666690190632
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,1536,0.06116977665159437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,1024,0.043826666143205434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2048,768,0.011979555918110741
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,1024,0.14772710535261366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,1024,0.013423999978436364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2048,512,0.009684444301658207
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,768,0.03682044479582045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,768,0.14102666907840306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,512,0.028192000256644353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,512,0.011863110793961419
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2048,256,0.007618666523032718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,512,0.1429191165500217
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,256,0.011092444260915121
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,256,0.14927377965715197
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2048,128,0.0069191112286514705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,3072,0.11656622091929118
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,128,0.01924533314175076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2048,64,0.006622222148709827
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,2560,0.09785866737365723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,128,0.13529245058695474
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,128,0.011241777903503843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,2048,32,0.006569777925809224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,64,0.0188391109307607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,32,0.01850844422976176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,2048,1536,0.1441777812110053
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1536,16384,0.14277422428131104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1536,65536,0.45467646916707355
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,2048,768,0.012416889270146688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,65536,0.29043645328945583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,16384,0.07712888717651367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1536,12288,0.10305333137512207
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,2048,256,0.02163822286658817
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,16384,0.591036425696479
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,12288,0.4443857669830322
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,16384,0.8476888868543836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,12288,0.6454870965745714
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,10240,0.37500445048014325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1536,8192,0.07680266433291964
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,10240,0.057051552666558154
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,8192,0.04490310947100321
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,10240,0.539356443617079
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1536,7168,0.06826844480302599
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,8192,0.30284354421827525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,8192,0.44051021999782985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1536,6144,0.05676622523201836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,7168,0.04059999850061204
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,7168,0.26558756828308105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,12288,0.06372533241907756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,7168,0.38934487766689724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,6144,0.035713778601752386
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1536,5120,0.042493333419164024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,6144,0.22435998916625977
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,6144,0.3409048981136746
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1536,4096,0.0365626679526435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,5120,0.03131555517514547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,5120,0.190701338979933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,5120,0.2889004548390706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1536,3584,0.030293332205878362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,4096,0.024159111910396155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,4096,0.23465956581963432
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1536,3072,0.0266248881816864
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,3584,0.13028177950117323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,3584,0.20604088571336535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,3072,0.10875999927520752
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,3072,0.02015733387735155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,3072,0.18314045005374482
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1536,2560,0.022820444570647344
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,65536,3.1980711619059243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,2560,0.01794666714138455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,2560,0.09188977877298991
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,2560,0.16026400195227727
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1536,2048,0.019679110911157396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1536,1536,0.014937778313954672
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1536,10240,0.09175911214616563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,65536,2.3530470530192056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,2048,0.016598222984208
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,2048,0.07445066505008273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,2048,0.1462657716539171
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1536,1024,0.011649777491887411
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,1536,0.05913688739140829
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,1536,0.013815999858909182
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1536,768,0.00996800015370051
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,1024,0.0120666664507654
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,1024,0.1462035576502482
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1536,512,0.007979555262459649
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,768,0.011397333608733283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,768,0.03326844506793552
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,768,0.14803022808498806
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,4096,0.1472302277882894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1536,256,0.007000000112586551
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,512,0.02638133366902669
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,512,0.0107360002067354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,512,0.1474835607740614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,256,0.0200311111079322
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1536,128,0.005912888795137405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,256,0.14841422769758436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,256,0.010009778042634329
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1536,64,0.0058417779703934984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,128,0.010064889159467485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,128,0.017879999346203275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,128,0.13896444108751085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1536,32,0.005910222315125995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,64,0.016905777984195285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,32,0.01687466601530711
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1024,65536,0.3425146738688151
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1024,16384,0.0937911139594184
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1536,1536,0.14435022407107884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1536,1024,0.040687110688951277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,16384,0.06507377492056952
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1024,12288,0.07179377476374309
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1536,3584,0.021956443786621094
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,16384,0.5697466532389323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,12288,0.05163910984992981
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,16384,0.7960337532891167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1024,10240,0.060434665944841176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,12288,0.41488001081678605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,12288,0.6091137992011176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,10240,0.045427554183536105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1024,8192,0.0492142207092709
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,65536,0.24170311292012533
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,10240,0.3484284347958035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1024,7168,0.04408355553944906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,10240,0.5166799757215712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,8192,0.281185785929362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,8192,0.4220026599036322
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,65536,2.172180387708876
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,7168,0.03299555513593886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,7168,0.2481155660417345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1024,5120,0.028861333926518757
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,7168,0.37207733260260684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,6144,0.029061334000693426
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,6144,0.2138915591769748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,5120,0.025536000728607178
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,5120,0.17772177855173746
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,5120,0.2771528826819526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,4096,0.14032088385687932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,4096,0.020651555723614164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,4096,0.22291289435492623
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1024,3584,0.02186044388347202
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,3584,0.01889244384235806
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,3584,0.12136177221934001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,65536,3.0468728807237415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1024,3072,0.01940799918439653
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,3584,0.19896354940202501
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,8192,0.03604711095492045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1024,2560,0.016965332958433364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,3072,0.017200888858901132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,3072,0.17376177840762666
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,3072,0.10158844126595391
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,2560,0.08535644743177627
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1024,6144,0.03525244527392917
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,2560,0.14900355868869358
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1024,2048,0.014469333820872836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,2560,0.015489778584904142
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1024,1536,0.011752888560295105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,2048,0.013629333012633853
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,2048,0.06988266441557142
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1024,1024,0.009298666483826106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,1536,0.012049777640236748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,1536,0.051960001389185585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,1536,0.14379555649227566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1024,768,0.007974222302436829
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1024,4096,0.02424088948302799
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,1024,0.03705244594150119
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,1024,0.13934222857157388
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1024,512,0.00666133314371109
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,768,0.030805332793129817
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,768,0.14834045039282903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1024,256,0.005632888939645555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,512,0.024679111109839544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,512,0.14968000517951116
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,256,0.018304889400800068
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1024,128,0.004622222234805425
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,256,0.13895289103190103
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,128,0.1331288946999444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,128,0.016134222348531086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1024,64,0.0042408886882993905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,128,0.009367999931176504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,1024,32,0.004545777622196409
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,64,0.015525332755512662
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,2048,0.14330489105648464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,1024,32,0.015455111861228943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,1024,6144,0.32335734367370605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,1024,0.011008000208271874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,768,65536,0.28151022063361275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,768,16384,0.07378044393327501
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,768,0.01036800030204985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,512,0.00997511131895913
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,16384,0.06442577971352471
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,65536,0.23171554671393502
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,1024,256,0.009355555805895064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,768,16384,0.5466648737589518
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,768,12288,0.40571199523078066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,12288,0.050642665889528066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,16384,0.7928782039218478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,12288,0.6062880092196995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,10240,0.04487555556827121
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,768,10240,0.3407884438832601
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,10240,0.5122248861524794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,768,12288,0.05469333463244968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,768,8192,0.27470045619540745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,768,7168,0.03210399879349603
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,8192,0.0351164440313975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,8192,0.41870665550231934
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,768,65536,2.110651652018229
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,768,6144,0.028304000695546467
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,768,7168,0.241650660832723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,7168,0.03195200032658047
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,768,10240,0.049934221638573535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,7168,0.36764176686604816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,6144,0.028110222684012517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,768,6144,0.20841422345903185
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,6144,0.3258177704281277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,768,4096,0.019283556275897555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,768,8192,0.03819288810094198
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,5120,0.025145777397685583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,768,3584,0.017242666747834947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,5120,0.2769546773698595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,4096,0.020561777883105807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,768,4096,0.136062224706014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,768,3072,0.01552088889810774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,4096,0.2221146689520942
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,768,3584,0.11787377463446723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,65536,3.0357750786675344
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,768,2560,0.013704000247849358
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,3072,0.1749253273010254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,768,2560,0.08367466926574707
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,2560,0.14875111314985487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,768,2048,0.011608888705571493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,2560,0.015616888801256815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,768,5120,0.024029332730505202
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,768,2048,0.06867021984524198
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,768,1536,0.009722666607962714
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,2048,0.1402382188373142
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,2048,0.013535999589496188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,768,5120,0.17364444997575548
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,1536,0.13986489507887098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,1536,0.012072000238630505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,768,1024,0.008083555433485243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,1024,0.01073155552148819
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,768,1024,0.03676889008945889
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,1024,0.1479697757297092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,768,768,0.006955555743641323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,3584,0.018972444865438674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,3584,0.199309335814582
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,768,512,0.005656888915432825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,768,0.010374222364690568
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,768,768,0.02978133492999607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,768,0.1478266716003418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,768,256,0.004898666507667965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,768,512,0.0236488895283805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,512,0.010015111002657149
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,512,0.14337510532803005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,768,128,0.004173333446184794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,768,256,0.017222222354676988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,256,0.009305777649084726
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,256,0.14067199495103624
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,768,64,0.003908444609906939
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,768,128,0.015431111057599386
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,768,32,0.004187555362780889
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,768,128,0.13784444332122803
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,768,64,0.014759999182489185
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,768,32,0.0144213338692983
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,512,65536,0.2111546728346083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,512,16384,0.055200000603993736
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,768,1536,0.051375110944112144
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,65536,0.1871537764867147
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,16384,0.05266666743490431
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,512,12288,0.04148711098564996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,512,16384,0.5253982014126247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,3072,0.017287999391555786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,12288,0.04155644443300035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,16384,0.7486969100104438
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,512,10240,0.033801777495278254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,512,12288,0.3899013201395671
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,10240,0.037456889947255455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,12288,0.5802391370137533
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,512,8192,0.026194666822751362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,512,10240,0.32781778441535103
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,10240,0.48889509836832684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,8192,0.028763555818133887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,512,7168,0.02231377859910329
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,512,8192,0.2645244333479139
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,8192,0.4020497798919678
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,512,6144,0.019578667150603402
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,7168,0.025802666942278545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,512,65536,2.024828381008572
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,512,7168,0.23283643192715117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,768,3072,0.09953066375520493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,512,5120,0.01667288939158122
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,7168,0.3566453456878662
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,6144,0.02318488889270359
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,512,6144,0.2002320024702284
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,512,4096,0.013768889009952545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,5120,0.021191999316215515
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,512,5120,0.16706755426194933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,5120,0.26392443974812824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,512,3584,0.012865777644846173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,768,128,0.009325332939624786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,4096,0.01680444512102339
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,512,4096,0.13115821944342718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,4096,0.21088088883293998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,512,3072,0.011565333439244164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,3584,0.015505777464972602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,3584,0.1883822282155355
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,512,2560,0.010585777461528778
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,65536,2.8921076456705728
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,512,3072,0.09536444478564793
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,3072,0.16767733626895479
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,512,2048,0.009143110778596666
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,512,2560,0.07985955476760864
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,2560,0.013447110851605734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,2560,0.14551111062367758
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,512,1536,0.0075795559419526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,512,2048,0.06520889202753703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,2048,0.01204355557759603
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,2048,0.14646844069163004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,512,1024,0.006268444574541516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,512,1536,0.04901599884033203
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,1536,0.14426222112443712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,1536,0.01070755554570092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,512,768,0.005596444424655702
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,512,1024,0.03438933359252082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,1024,0.009941333697901832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,6144,0.3116942246754964
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,1024,0.13959022363026938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,512,512,0.0047493332790003884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,512,768,0.028689778513378564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,768,0.14639200104607478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,512,256,0.0042204442951414315
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,512,512,0.023003555006451074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,512,0.1435599989361233
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,512,0.009011555876996782
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,512,128,0.0038248888320393036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,512,256,0.016520889268981088
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,512,3584,0.11287821663750543
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,512,128,0.0144195556640625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,512,64,0.003555555724435382
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,3072,0.014447111222479077
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,128,0.008533333738644918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,128,0.1403991116417779
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,512,32,0.003569777641031477
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,512,64,0.013804444836245643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,512,32,0.013536888692114087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,256,16384,0.03917422228389316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,256,65536,0.1241048839357164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,65536,0.18620799647437203
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,16384,0.052271998590893216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,256,12288,0.02942044536272685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,768,0.009349333743254343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,256,16384,0.49663909276326496
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,12288,0.04102222124735514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,512,256,0.13972800307803684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,16384,0.7500221994188098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,256,10240,0.021902221772405837
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,256,12288,0.37556176715426975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,10240,0.03698933455679152
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,256,8192,0.017322666115230985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,12288,0.5747484631008571
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,256,10240,0.31483732329474556
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,256,7168,0.015488889482286243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,10240,0.4850471284654405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,256,8192,0.25448621643914116
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,8192,0.3990328841739231
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,7168,0.024754666619830664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,256,6144,0.014005333185195923
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,256,7168,0.22342665990193686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,7168,0.3519555462731256
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,256,65536,1.9473084343804254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,6144,0.022944889134830896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,256,5120,0.012798222402731577
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,256,6144,0.19290044572618273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,6144,0.3112257851494683
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,256,4096,0.010196444061067369
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,5120,0.021095110310448542
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,512,256,0.008611555728647444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,256,5120,0.16087555885314941
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,5120,0.2639839914109972
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,4096,0.016895999511082966
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,256,4096,0.1255120038986206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,4096,0.20732266373104521
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,256,3072,0.008759111166000366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,3584,0.015726221932305228
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,256,3584,0.10781511333253647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,3584,0.19010667006174722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,256,2560,0.00794400026400884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,256,3072,0.09095644288592869
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,3072,0.16359644465976292
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,256,2048,0.006907555378145642
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,8192,0.027679999669392902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,65536,2.873986562093099
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,256,2560,0.07553333044052124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,256,2048,0.061195552349090576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,2048,0.01204355557759603
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,256,1536,0.0063271112740039825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,256,1536,0.04619910981920031
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,1536,0.142593781153361
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,256,1024,0.005418666534953647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,1536,0.010724444356229572
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,256,1024,0.032646222247017756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,256,3584,0.009693333672152625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,256,768,0.004896000027656555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,1024,0.00999111102686988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,256,768,0.027441778116756018
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,768,0.14170400301615396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,3072,0.014467555615637036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,768,0.009317333499590555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,2560,0.013403555585278405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,256,512,0.02164177762137519
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,256,256,0.0035253332720862497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,512,0.14514577388763428
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,512,0.008998221821255153
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,256,128,0.0035111109415690103
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,256,256,0.01593244406912062
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,256,0.008392888638708327
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,256,128,0.013468444347381592
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,256,64,0.0031840000301599503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,256,128,0.008654221892356873
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,128,0.13744000593821207
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,256,32,0.0031511110977994073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,256,64,0.013136000268989138
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,256,32,0.012726222475369772
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,1024,0.13962133725484213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,65536,0.18615378273857963
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,256,512,0.004206222378545337
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,128,16384,0.029948443174362183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,2560,0.14867022302415636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,2048,0.1422871086332533
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,16384,0.05362133185068766
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,256,256,0.14704355928632948
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,128,12288,0.023195554812749226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,128,16384,0.4861031108432346
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,128,10240,0.016495111915800307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,12288,0.04188444548183017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,128,12288,0.3672746552361383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,128,65536,0.09500088956620957
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,16384,0.7478453318277994
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,128,8192,0.013532444834709167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,10240,0.03797244363360935
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,128,10240,0.30815378824869794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,128,7168,0.012430222498046027
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,8192,0.029153777493370905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,10240,0.48571379979451496
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,8192,0.39926934242248535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,128,6144,0.011061333119869232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,128,7168,0.2190177838007609
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,7168,0.02597066594494714
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,128,65536,1.9063671959771051
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,128,5120,0.010035555395815108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,7168,0.3544062243567572
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,6144,0.02349866595533159
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,128,6144,0.18849155637953016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,128,4096,0.008330666356616551
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,5120,0.021485333641370136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,128,5120,0.1577893363104926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,5120,0.26259732246398926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,128,3584,0.007695111135641734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,128,4096,0.12343111303117539
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,128,3584,0.10525866349538167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,12288,0.5725555419921875
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,3584,0.18933155801561144
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,128,3072,0.007300444775157505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,65536,2.8660248650444875
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,128,8192,0.2485884560479058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,3584,0.016175111134847004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,128,2560,0.006614222294754452
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,3072,0.16294222407870823
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,128,3072,0.08876799874835545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,128,2048,0.005910222315125995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,128,2560,0.07469066646364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,2560,0.013723555538389417
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,128,2048,0.060657779375712075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,2048,0.14482755131191677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,2048,0.012118221984969245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,6144,0.30991466840108234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,128,1536,0.04555199874771965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,1536,0.1475600004196167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,128,1024,0.004510222209824456
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,1536,0.010819555156760745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,4096,0.01755466726091173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,128,768,0.003954666770166821
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,128,1024,0.031678222947650485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,1024,0.14825956026713052
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,1024,0.00997866690158844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,128,512,0.00360000009338061
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,128,768,0.026723555392689172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,768,0.009368889033794403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,128,512,0.02124533388349745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,512,0.008983111215962304
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,128,256,0.0031928889867332247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,512,0.14598933855692545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,3072,0.01496266656451755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,128,256,0.01517599986659156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,256,0.14275999863942465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,2560,0.1426186694039239
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,128,128,0.013060444758998023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,128,1536,0.005439110928111606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,128,64,0.0028320000403457214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,128,0.008485332959228093
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,128,32,0.0031439999325407874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,128,64,0.01239111108912362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,128,32,0.012459555433856117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,4096,0.20793510807885063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,64,65536,0.0882906648847792
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,64,16384,0.028485334581798975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,64,12288,0.019522666931152344
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,768,0.14608177873823378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,64,10240,0.015646222564909194
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,128,128,0.0028773333049482773
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,64,12288,0.36697954601711696
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,64,16384,0.4854239887661404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,64,10240,0.3065688874986437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,64,7168,0.011355555719799466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,64,6144,0.010306666294733683
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,64,8192,0.24846132596333823
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,64,7168,0.21702844566769072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,64,5120,0.009000000026490953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,64,4096,0.007861333588759104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,64,3584,0.007269333634111617
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,64,5120,0.15610488255818686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,64,4096,0.12343022558424209
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,64,3072,0.006620444357395172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,64,2560,0.0059262220230367445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,64,3072,0.08825421995586819
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,64,2560,0.07275999916924371
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,64,2048,0.05892088678148058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,64,8192,0.012744888663291931
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,64,1536,0.004568888909286923
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,64,1024,0.0038737778862317405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,64,1536,0.045497778389188975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,64,768,0.0038417776425679526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,64,6144,0.18838044007619223
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,64,768,0.02647288971477085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,64,512,0.020960888928837244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,64,3584,0.10537244213951959
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,64,256,0.0033200000309281876
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,64,128,0.002862222285734283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,64,256,0.015136000182893542
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,64,2048,0.005584888988071018
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,64,128,0.012793777717484368
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,64,32,0.002809777855873108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,64,64,0.012411555482281579
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,64,65536,1.9049243927001953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4096,128,256,0.008356444537639618
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,64,32,0.012097777591811286
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,32,16384,0.026719111535284255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,32,12288,0.016484444340070088
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,32,65536,0.08937778075536092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,64,1024,0.031535999642478094
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,32,10240,0.011749332977665795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,32,8192,0.010388444695207808
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,64,512,0.003539555602603488
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,32,16384,0.48571819729275173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,32,7168,0.009700444837411245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,32,6144,0.008641777767075432
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,32,10240,0.3065191109975179
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,32,8192,0.24893776575724283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,32,5120,0.011769777370823754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,32,4096,0.010042666561073726
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,32,7168,0.21860710779825845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,32,6144,0.1882115602493286
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,32,3584,0.008650666309727563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,32,5120,0.15755555364820692
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,32,4096,0.12228000164031982
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,32,2560,0.007632888853549957
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,32,3584,0.10560444328520034
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,32,3072,0.08917422427071466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,32,2048,0.0052560000783867305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,32,1536,0.00488711107108328
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,32,2560,0.07337422503365411
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,32,1024,0.0041724443435668945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,32,1536,0.044663998815748424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,32,768,0.0038124442928367188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,32,1024,0.031472888257768415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,32,512,0.003569777641031477
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,32,768,0.026388444834285315
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,32,12288,0.36675911479526097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,32,256,0.0031519999934567344
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,32,128,0.0031395554542541504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,32,512,0.020929776959949072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,32,64,0.0028106667515304354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,32,256,0.015136889285511441
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,32,32,0.002830222249031067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,32,128,0.012729778057999082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,32,64,0.012094222009181976
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,64,64,0.0028488888508743713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4096,32,3072,0.007976888782448238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,32,32,0.012069332930776807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,32,2048,0.06026222308476766
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4096,32,65536,1.9058187272813585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,16384,0.7977253595987955
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,65536,12288,1.8777376810709636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,65536,16384,2.467386669582791
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,12288,2.1656311882866754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,12288,0.6214666896396214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,16384,3.1091456943088107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,65536,10240,1.6492364671495225
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,10240,0.5287831094529893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,12288,2.092611524793837
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,16384,2.773771497938368
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,10240,1.7729742262098525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,8192,0.4288960032992893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,65536,8192,1.267157342698839
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,65536,7168,1.1587093141343858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,8192,1.3994568718804254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,10240,1.8432896931966145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,7168,0.3740364445580377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,7168,1.1911546919080946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,8192,1.4101955625745985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,6144,0.32592977417839897
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,65536,6144,1.0302409066094291
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,65536,5120,0.7521875699361166
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,6144,1.0313048892550998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,7168,1.2296586566501193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,5120,0.3396871089935303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,65536,4096,0.6136488914489746
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,5120,0.8832417594061958
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,6144,1.020714653862847
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,4096,0.6852515538533529
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,4096,0.2366408771938748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,5120,0.8293626573350695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,65536,3584,0.5197902255588108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,3584,0.2130204439163208
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,3584,0.6028578016493056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,4096,0.6986239751180013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,65536,3072,0.49960533777872723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,3072,0.18802311685350206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,65536,2560,0.36525334252251523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,3584,0.6220257547166612
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,3072,0.5583217938741049
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,2560,0.16301600138346353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,2560,0.46131642659505206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,65536,2048,0.2982649008433024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,3072,0.5458853509691026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,2048,0.1382924450768365
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,2048,0.38644001219007706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,65536,1536,0.2411733203464084
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,2048,0.3869484530554877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,65536,1024,0.17268710666232637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,1536,0.3414444393581814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,1536,0.3138977686564128
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,65536,768,0.13773511515723333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,1024,0.08807733323838975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,1024,0.2547617753346761
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,65536,512,0.11173510551452637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,768,0.07723822196324666
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,768,0.2101271152496338
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,512,0.17112978299458823
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,512,0.06900178061591254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,512,0.14926666683620876
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,65536,256,0.06903821892208524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,256,0.14771289295620388
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,256,0.13137511412302652
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,256,0.06242577897177803
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,65536,128,0.05747999747594198
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,2560,0.5071040259467231
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,65536,128,0.13919466071658665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,128,0.12412444750467937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,1536,0.11132621765136719
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,65536,128,0.05958844555748833
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,65536,32,0.050107555256949536
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,1024,0.22722933027479383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,64,0.12560533152686224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,32,0.12602310710483128
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,65536,768,0.17576621638404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,16384,16384,0.6464346779717339
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,65536,64,0.04833066794607374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,16384,0.9027840296427408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,65536,0.9696577919854058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,16384,0.21448532740275064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,16384,12288,0.4585288895501031
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,12288,0.1640239953994751
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,16384,0.9370231628417969
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,16384,65536,2.672540452745226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,12288,0.8530062039693197
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,12288,0.6910275353325738
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,16384,10240,0.3581111166212294
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,10240,0.14228088325924343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,16384,8192,0.3017866611480713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,10240,0.5664755503336588
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,8192,0.12324710687001546
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,16384,7168,0.27964888678656685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,10240,0.5989617771572536
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,8192,0.477466689215766
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,7168,0.10360444254345363
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,7168,0.40052000681559247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,16384,6144,0.25642845365736217
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,7168,0.4600515365600586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,6144,0.09267999728520711
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,16384,5120,0.2005235619015164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,65536,4.375712076822917
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,5120,0.28620976871914333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,8192,0.47080442640516496
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,16384,4096,0.14435822433895534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,5120,0.07943999767303467
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,5120,0.3386062251196967
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,4096,0.06543733014000787
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,4096,0.2149653302298652
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,4096,0.26526133219401044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,65536,3.7816204494900174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,6144,0.3221324549780952
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,3584,0.1893528832329644
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4096,128,128,0.1373386647966173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,3584,0.23451378610399035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,6144,0.35989954736497665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,3584,0.05841066439946493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,16384,2560,0.10812444157070583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,3072,0.05193510982725355
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,2560,0.046000887950261436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,2560,0.1476115518146091
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,16384,2048,0.08891822232140435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,2048,0.11967999405331081
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,2048,0.15732533401913112
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,2048,0.0396035545402103
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,16384,1536,0.06245333618587918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,16384,3584,0.1541608836915758
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,16384,1024,0.04344711038801405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,1536,0.14868266052669948
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,1536,0.03254844413863288
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,1536,0.09534400039248997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,16384,768,0.03459377752410041
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,1024,0.0728435582584805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,1024,0.1477875577078925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,16384,3072,0.13029778003692627
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,768,0.059531556235419385
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,16384,512,0.02522666586769952
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,768,0.024524443679385718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,768,0.14543022049797907
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,3072,0.1665324502521091
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,16384,256,0.0196426659822464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,512,0.04877777894337972
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,512,0.02236266599761115
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,16384,128,0.018535999788178336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,2560,0.1838302214940389
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,256,0.14637333816952175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,128,0.04007555709944831
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,128,0.020119110743204754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,16384,64,0.015830222103330824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,16384,32,0.01581333412064446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,64,0.03863999909824795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,32,0.03874399926927354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,1024,0.02697511182890998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,3072,0.21001244915856254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,512,0.145414219962226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,16384,256,0.04283644424544441
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,16384,256,0.020103999310069613
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,16384,128,0.13317510816786024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,12288,16384,0.47393157747056747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,16384,0.16977955235375297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,12288,12288,0.3851795461442735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,16384,0.7278319994608561
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,65536,0.7499901983473035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,12288,0.1315973334842258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,12288,65536,2.0037814246283636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,16384,0.7950915760464139
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,12288,0.5302337540520562
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,12288,10240,0.3913502163357205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,10240,0.11390311188167995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,12288,8192,0.2238044473859999
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,10240,0.4447004530164931
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,12288,0.6205315589904785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,8192,0.08989689085218643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,12288,7168,0.1956284443537394
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,8192,0.4864782227410211
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,8192,0.4128444459703233
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,7168,0.08103022310468885
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,7168,0.3083137671152751
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,12288,6144,0.2314986652798123
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,7168,0.3703502284155951
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,12288,5120,0.14904622236887613
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,6144,0.26155821482340497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,6144,0.07191466622882418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,6144,0.3161911169687907
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,5120,0.06304444207085504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,12288,4096,0.1347244448131985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,65536,3.336990144517687
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,5120,0.2188711166381836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,5120,0.2888222270541721
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,12288,3584,0.12200178040398492
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,4096,0.052299555804994374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,4096,0.1763368844985962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,4096,0.22690134578280977
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,12288,3072,0.10308088858922322
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,3584,0.1558657752143012
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,3584,0.046075556013319224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,3584,0.21159556176927355
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,10240,0.5260533226860894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,12288,2560,0.07316533062193128
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,3072,0.040940443674723305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,3072,0.18771110640631783
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,12288,2048,0.06011111206478543
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,2560,0.12608355946011013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,2560,0.16352355480194092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,12288,1536,0.05190133386188083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,2048,0.10139199760225083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,2048,0.03129422333505418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,1536,0.07963289154900445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,1536,0.025477333201302424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,12288,1024,0.03424177898301019
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,65536,3.1905475192599826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,1024,0.13983910613589817
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,12288,768,0.031664000617133245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,1024,0.02252799934811062
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,768,0.04871288935343424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,768,0.1450551086001926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,12288,512,0.024344000551435683
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,768,0.01891288823551602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,3072,0.13764354917738172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,512,0.04047466648949517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,12288,256,0.016957332690556843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,512,0.01741155485312144
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,512,0.14076799816555446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,2560,0.03576888971858554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,12288,128,0.013760888742076026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,256,0.01613066593805949
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,256,0.14631466070810953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,256,0.034155554241604276
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,1536,0.14588622252146402
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,12288,64,0.012695111334323883
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,128,0.032243556446499295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,12288,32,0.013055110971132914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,128,0.13388178083631727
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,12288,128,0.01610933409796821
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,64,0.03156000044610765
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,32,0.031167109807332356
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,12288,1024,0.060623109340667725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,10240,16384,0.4298115571339925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,16384,0.6470470958285862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,12288,2048,0.14641066392262778
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,65536,0.6302231152852377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,16384,0.14945067299736872
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,10240,12288,0.2999324533674452
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,10240,65536,1.6217315461900499
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,12288,0.1142702235115899
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,10240,10240,0.25884887907240123
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,16384,0.7382071283128527
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,10240,0.09743911027908325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,12288,0.5491537517971462
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,10240,8192,0.22945777575174967
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,10240,0.4581431282891168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,10240,0.46620623270670575
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,8192,0.07914577590094672
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,8192,0.3285866578420003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,10240,7168,0.16849066151512992
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,8192,0.37395644187927246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,10240,6144,0.15060799651675755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,7168,0.07073422273000081
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,7168,0.2739368809594048
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,10240,5120,0.12163289388020833
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,6144,0.06240000327428182
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,6144,0.23114665349324545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,65536,2.893096923828125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,12288,0.5744577513800727
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,5120,0.19473777876959908
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,5120,0.24817066722446016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,4096,0.15699288580152723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,4096,0.0437395539548662
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,4096,0.20931555165184867
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,10240,3584,0.09458400143517388
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,3584,0.1402719948026869
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,3584,0.038098666403028704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,3584,0.1894986629486084
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,10240,3072,0.08223555485407512
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,65536,2.8284960852728953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,3072,0.17060444090101454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,3072,0.12292444705963135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,10240,2560,0.06972889105478923
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,7168,0.34051556057400173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,10240,2048,0.05755822526084053
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,2560,0.15080977810753718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,2560,0.030942221482594807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,2560,0.10597511132558186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,6144,0.2966639995574951
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,10240,1536,0.0443120002746582
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,5120,0.05281688769658407
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,2048,0.026471111509535048
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,10240,4096,0.1160222159491645
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,10240,1024,0.031162665949927434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,1536,0.07054489188724093
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,1536,0.1461928950415717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,1536,0.02148088812828064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,10240,768,0.02295911146534814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,1024,0.052417778306537204
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,1024,0.018274666534529794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,768,0.04375466704368591
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,768,0.1452853282292684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,768,0.01662400033738878
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,512,0.034661332766215004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,512,0.14501778284708658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,10240,256,0.013784888717863293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,256,0.03158933255407546
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,3072,0.034020443757375084
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,256,0.014298665854665967
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,10240,128,0.011653333074516721
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,2048,0.08855022324456109
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,2048,0.14668888515896267
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,128,0.014153778553009033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,10240,64,0.010326222413116032
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,1024,0.14865422248840332
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,10240,32,0.010599999792046018
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,64,0.02849422229660882
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,32,0.02812533246146308
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,10240,512,0.01574488812022739
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,256,0.1428826649983724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,10240,128,0.028593778610229492
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,10240,128,0.13803022437625462
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,10240,512,0.018779555956522625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,8192,16384,0.31835820939805776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,8192,65536,1.2300959693060982
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,16384,0.12189600202772354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,8192,12288,0.26452088356018066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,16384,0.5616835488213433
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,12288,0.09302666452195908
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,8192,10240,0.2231760025024414
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,12288,0.43332621786329484
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,16384,0.6485395431518555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,12288,0.4978124300638835
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,10240,0.08065599865383573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,8192,8192,0.1829368935690986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,10240,0.35383110576205784
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,65536,0.5066711107889811
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,8192,7168,0.1508400042851766
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,10240,0.4506310886806912
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,8192,0.27422576480441624
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,8192,0.06580533583958943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,65536,2.5320470598008895
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,8192,0.34085954560173887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,7168,0.05860888957977295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,7168,0.23817867702907988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,8192,5120,0.11579022142622207
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,6144,0.051573332813051015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,6144,0.26799554295010036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,5120,0.1696079969406128
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,5120,0.04474400149451362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,8192,4096,0.09105422099431355
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,5120,0.2274728880988227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,8192,3584,0.07846222321192424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,4096,0.03777599996990628
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,4096,0.18944888644748262
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,4096,0.13673333326975504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,65536,2.51875368754069
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,8192,3072,0.05970222420162625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,3584,0.12390932771894668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,3584,0.17412977748446992
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,8192,2560,0.0569288863076104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,3072,0.02858933475282457
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,3072,0.1559048891067505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,3072,0.11012444231245254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,8192,2048,0.04287022352218628
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,2560,0.09370044204923843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,8192,6144,0.11678755283355713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,2560,0.14457066853841147
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,2560,0.026187555657492742
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,8192,1536,0.03380889031622145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,2048,0.07811288701163398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,2048,0.022304000125990972
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,2048,0.14980267153845891
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,7168,0.30599554379781085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,1536,0.061554663711123996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,8192,1024,0.025145777397685583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,1536,0.14333866702185735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,8192,768,0.01940444442960951
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,1024,0.046257777346505054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,1024,0.14580710728963217
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,8192,512,0.015056000815497505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,768,0.0396364430586497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,768,0.14672266112433538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,768,0.014739554789331226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,8192,256,0.01237777786122428
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,512,0.013857777747843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,512,0.14485066466861302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,256,0.02608088817861345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,3584,0.03161955542034573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,256,0.1455813381406996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,8192,128,0.01037599974208408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,128,0.012795555922720166
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,128,0.02506133251720005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,8192,64,0.009663999908500248
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,8192,128,0.1850551101896498
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,8192,32,0.01000711073478063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,64,0.025032000409232244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,32,0.024940444363488093
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,6144,0.20621600415971544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,1536,0.019260444574885897
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,1024,0.016183111402723525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,7168,16384,0.3004293441772461
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,8192,512,0.03183111217286851
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,8192,256,0.01273155543539259
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,65536,0.4643840259975857
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,16384,0.11010044150882298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,16384,0.5133955743577745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,7168,65536,1.197797351413303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,7168,12288,0.241592009862264
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,12288,0.0852924452887641
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,7168,10240,0.2191013362672594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,12288,0.48024712668524844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,16384,0.636864874098036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,12288,0.4873618019951715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,7168,8192,0.16041689448886448
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,10240,0.07311289178000556
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,10240,0.32765155368381077
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,8192,0.059373332394493945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,8192,0.25442579057481557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,10240,0.44698577457004124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,7168,7168,0.1346400048997667
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,8192,0.33453954590691465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,7168,6144,0.11205955346425374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,7168,0.053582221269607544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,7168,0.2235297891828749
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,7168,0.29781421025594074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,65536,2.3463022443983292
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,6144,0.047037333250045776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,6144,0.18892178270551893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,7168,5120,0.10231110784742568
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,6144,0.2611582279205322
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,7168,4096,0.08202577961815728
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,5120,0.15695022212134466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,5120,0.22776532173156738
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,7168,3584,0.06183289157019722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,4096,0.12812977366977268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,4096,0.1866906616422865
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,4096,0.03408266769515143
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,7168,3072,0.05925955375035604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,3584,0.11373599370320638
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,3584,0.1709271139568753
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,3584,0.030332446098327637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,3072,0.1555511156717936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,7168,2560,0.051727109485202365
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,3072,0.02782400117980109
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,7168,2048,0.03766577773623996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,2560,0.08842488792207505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,65536,2.443779627482096
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,7168,1536,0.0302746660179562
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,2048,0.07242933246824476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,2048,0.14661778344048396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,2048,0.021032000581423443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,7168,1024,0.02128444446457757
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,1536,0.05683822102016873
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,1536,0.0177511109246148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,1024,0.04328533344798618
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,7168,768,0.01742488808102078
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,1024,0.015060444672902426
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,1024,0.14578399393293592
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,5120,0.04139822059207492
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,768,0.03646577729119195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,7168,512,0.014441778262456259
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,768,0.014007111390431723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,512,0.029110223054885864
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,512,0.14483200179206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,7168,256,0.010404444403118556
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,256,0.025407110651334126
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,256,0.14688889185587564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,2560,0.02499288817246755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,7168,128,0.009296889106432596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,256,0.012167111039161682
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,128,0.023415111833148535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,128,0.13966400093502468
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,7168,64,0.008296888735559251
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,128,0.012156444291273752
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,7168,32,0.008647110727098253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,1536,0.1441617806752523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,64,0.0233608881632487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,32,0.023028444912698533
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,768,0.14669866032070583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,7168,512,0.013007111019558378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,65536,0.4090462260776096
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,7168,3072,0.10171733299891154
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,7168,2560,0.1420933273103502
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,6144,65536,0.9182266659206815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,16384,0.09625333547592163
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,6144,12288,0.18476177586449516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,16384,0.47402132882012266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,12288,0.07443644603093465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,12288,0.3827822208404541
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,16384,0.6474471092224121
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,12288,0.4606622060139974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,10240,0.2999262279934353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,10240,0.06538133488761054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,6144,8192,0.1295644442240397
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,10240,0.40329421891106504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,6144,16384,0.24287732442220053
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,8192,0.05193600058555603
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,8192,0.2356977727678087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,65536,2.0533777872721353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,8192,0.32148533397250706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,6144,6144,0.09924533632066514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,7168,0.20425332917107475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,7168,0.04716533422470093
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,7168,0.2839404477013482
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,6144,0.17464710606469047
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,6144,0.04177422324816386
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,6144,10240,0.1459724505742391
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,6144,0.25777777036031085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,6144,4096,0.06653777758280437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,5120,0.21012532711029053
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,5120,0.036976890431510076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,65536,2.3002692328559027
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,4096,0.11979556083679199
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,4096,0.03071199854214986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,4096,0.17802577548556855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,6144,3584,0.0603662199444241
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,6144,3072,0.05365066727002462
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,3584,0.02770400047302246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,3584,0.16204267077975804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,6144,7168,0.11639732784695095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,3072,0.09204889006084865
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,3072,0.14749956130981445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,3072,0.025248888466093276
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,6144,2560,0.04125600059827169
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,6144,2048,0.03268266717592875
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,2560,0.022486221459176805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,2560,0.1431422233581543
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,2560,0.08088977469338311
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,6144,1536,0.02606044378545549
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,2048,0.06622666782803006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,2048,0.01937866707642873
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,2048,0.14796622594197592
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,6144,1024,0.019796444310082328
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,1536,0.05312622255749173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,1536,0.016936888297398884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,1536,0.15046932962205675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,6144,768,0.016158221496476066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,1024,0.040186666780047946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,1024,0.14863377147250706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,768,0.0337084432442983
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,6144,512,0.012255111502276527
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,768,0.1473671065436469
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,768,0.01312355531586541
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,3584,0.10659022463692559
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,512,0.02676533328162299
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,512,0.1471173365910848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,6144,256,0.0120328888297081
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,256,0.02288444505797492
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,256,0.01148799972401725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,6144,128,0.010647111468844943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,6144,5120,0.0720506641599867
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,128,0.02199199961291419
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,128,0.14509422249264187
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,6144,64,0.009660444325870937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,6144,32,0.010063110954231685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,128,0.011742221812407175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,64,0.02196177840232849
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,5120,0.1447262234157986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,6144,32,0.021651556094487507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,1024,0.014429334137174817
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,5120,16384,0.2238453229268392
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,6144,512,0.012260444462299347
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,65536,0.35625423325432676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,5120,65536,0.7706089019775391
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,16384,0.43886754247877335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,16384,0.08665066957473755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,6144,256,0.14995288848876953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,5120,12288,0.18306044737497965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,5120,10240,0.13909510771433511
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,12288,0.06565422481960721
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,16384,0.5819377899169922
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,12288,0.35663644472757977
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,10240,0.05717066923777262
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,5120,8192,0.11368710464901394
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,10240,0.3065688874986437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,5120,7168,0.09022044473224217
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,8192,0.046035554673936635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,10240,0.37040087911817765
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,8192,0.2204160028033786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,8192,0.29741867383321124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,7168,0.04160799913936191
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,5120,6144,0.07208444674809773
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,7168,0.1920622189839681
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,7168,0.26774220996432835
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,5120,5120,0.06595644685957167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,6144,0.03779733180999756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,65536,1.8727092742919922
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,6144,0.15950577788882786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,6144,0.2290817896525065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,5120,4096,0.05506933397716946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,5120,0.03240888979699876
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,5120,0.19797333081563315
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,4096,0.027217777238951787
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,4096,0.1681377755271064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,3584,0.09789866871303982
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,3584,0.15308445029788548
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,3584,0.024682665864626568
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,5120,3072,0.04052088989151849
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,12288,0.441011561287774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,3072,0.0222062220176061
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,3072,0.08516889148288304
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,2560,0.14062933127085367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,2560,0.07156266768773396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,5120,2048,0.02696711156103346
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,2560,0.020213334096802604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,2048,0.06149333053165012
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,2048,0.14023999373118082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,5120,0.1339635584089491
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,2048,0.0169804443915685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,5120,1536,0.022249778111775715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,5120,3584,0.04334133201175266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,1536,0.01512888901763492
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,1536,0.04875822199715508
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,1536,0.14499910672505698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,5120,1024,0.01626755628320906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,1024,0.03684355484114753
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,1024,0.14315644900004068
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,1024,0.012952889005343119
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,5120,2560,0.03793777690993415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,3072,0.1475600004196167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,5120,512,0.010553778045707278
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,768,0.03155022197299533
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,768,0.14521866374545628
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,512,0.024754666619830664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,5120,256,0.007970666719807519
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,512,0.14352088504367405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,512,0.011371555427710215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,256,0.02132533325089349
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,5120,128,0.007120889094140794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,256,0.01072533345884747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,256,0.14355644914839003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,128,0.020640000700950623
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,128,0.01072266697883606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,128,0.13876267274220785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,5120,64,0.006321777900060018
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,5120,32,0.006606222026877933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,64,0.020242666204770405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,32,0.019991111424234178
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,5120,768,0.012921777864297232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,5120,65536,2.149854236178928
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,5120,768,0.012060444388124677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,4096,65536,0.5762462086147732
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,4096,16384,0.17151378260718453
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,4096,12288,0.12860000133514404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,16384,0.07225600216123793
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,65536,0.30573421054416233
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,16384,0.40020354588826496
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,12288,0.055255111720826894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,4096,10240,0.11184888415866429
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,5120,4096,0.1095128920343187
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,12288,0.30067290200127494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,16384,0.5536319944593642
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,4096,8192,0.0881226658821106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,10240,0.04841244551870558
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,10240,0.2548675537109375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,4096,7168,0.0763875577184889
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,8192,0.0392213331328498
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,8192,0.19827732774946424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,8192,0.2771013312869602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,4096,6144,0.06864088773727417
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,65536,1.6534658008151586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,7168,0.17173155148824057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,7168,0.24828177028232148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,4096,5120,0.056741330358717174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,6144,0.03193333413865831
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,6144,0.14541510740915933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,6144,0.2170497841305203
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,4096,4096,0.0458586679564582
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,5120,0.028661333852344092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,5120,0.12172178427378337
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,5120,0.1845608949661255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,4096,3584,0.04001511136690775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,4096,0.09968800014919704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,4096,0.023642667465739783
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,4096,0.1597786611980862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,65536,1.9942745632595484
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,4096,3072,0.033972445461485125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,12288,0.411589331097073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,3584,0.08828444613350762
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,3584,0.144813338915507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,4096,2560,0.028909332222408716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,3072,0.07668444183137682
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,3072,0.15076355139414468
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,4096,2048,0.02446222139729394
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,10240,0.3477333386739095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,2560,0.06542222367392646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,2560,0.1427324480480618
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,2560,0.017649778061442904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,4096,1536,0.019861333900027804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,2048,0.05388444330957201
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,2048,0.14568889141082764
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,4096,1024,0.01481066644191742
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,1536,0.04359822140799629
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,1536,0.14074577225579157
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,1536,0.013780444032616086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,4096,768,0.012275555895434486
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,1024,0.03333510955174764
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,1024,0.012093333734406365
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,1024,0.14660444524553087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,4096,512,0.009349333743254343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,768,0.027922666735119287
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,768,0.011053333679835001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,768,0.14546399646335179
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,4096,256,0.007315555380450354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,512,0.02288888891537984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,512,0.010677333507272931
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,512,0.15100622177124023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,256,0.019664888580640156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,256,0.1472124391131931
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,4096,128,0.006937777830494775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,3584,0.021729777256647747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,3072,0.019488000207477145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,128,0.018939556346999276
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,128,0.00998222248421775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,4096,64,0.007225777539942
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,4096,128,0.13897866672939724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,4096,32,0.007527111305130853
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,64,0.018552889426549275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,4096,32,0.01824711097611321
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,7168,0.036377777655919395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3584,16384,0.152712000740899
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,65536,0.2787253326839871
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3584,65536,0.6505324575636122
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,16384,0.08249777555465698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,16384,0.3789466751946344
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,256,0.009959110783206092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3584,12288,0.12933511204189724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,16384,0.5313075383504232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,12288,0.28550932142469615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3584,10240,0.10767289002736409
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,12288,0.4095253414577908
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,10240,0.23881244659423828
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,65536,1.590816921657986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,10240,0.04646844334072537
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3584,8192,0.07901422182718913
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,10240,0.3404640091790094
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3584,7168,0.07526044713126288
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,8192,0.18895910845862496
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,8192,0.03798222210672166
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,8192,0.277566221025255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3584,6144,0.06492977672153048
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,7168,0.03443377878930833
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,7168,0.16291644838121203
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,65536,1.9794186486138239
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,6144,0.1387057834201389
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,6144,0.031207998593648274
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3584,5120,0.053096890449523926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,6144,0.21405600176917183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,12288,0.05778488847944471
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3584,4096,0.042734222279654614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,5120,0.02760533326201969
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,5120,0.18510222434997559
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,5120,0.11659467220306396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,4096,0.02304977840847439
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,4096,0.09563555320103963
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3584,3072,0.033534222178988986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,3584,0.021002666817771062
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,3584,0.08490400181876288
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,3584,0.146741337246365
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,3072,0.07537866963280572
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,3072,0.018840889135996502
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,3072,0.13957777288224962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,2560,0.06406666835149129
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,2560,0.017216889394654166
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,2560,0.15130489402347141
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3584,2048,0.024347555306222703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,2048,0.051741331815719604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,7168,0.24638220998975965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,2048,0.150346663263109
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3584,1536,0.01964000033007728
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,1536,0.01348533315791024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,1536,0.04268711143069797
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,1536,0.14848355452219644
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3584,1024,0.014931556251313953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,4096,2048,0.015616888801256815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,1024,0.011696000066068439
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3584,3584,0.039220445685916476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3584,768,0.012675555215941535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,768,0.02686400049262577
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,768,0.14202933841281468
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3584,512,0.009341333475377824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3584,2560,0.02916533417171902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,512,0.021926222576035395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,512,0.13901155524783665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3584,256,0.007684444387753804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,512,0.01034311122364468
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,256,0.01886133352915446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,256,0.14178755548265246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3584,128,0.006595555692911148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,2048,0.015092444088723926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,128,0.017869333426157635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,128,0.1331511073642307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3584,64,0.006569777925809224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3584,32,0.00756888919406467
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,64,0.017888888716697693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,32,0.01756622228357527
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3584,1024,0.03150666753451029
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,4096,0.15845599439409044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,768,0.011055111057228513
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3072,65536,0.46341779496934676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3072,16384,0.14013955328199598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,65536,0.24427911970350477
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,16384,0.067958222495185
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,256,0.009696889254781935
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3072,12288,0.10831466648313735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,16384,0.3755679925282796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3584,128,0.00996800015370051
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,16384,0.4972817632887099
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,12288,0.27171288596259224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3072,10240,0.09669688675138687
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,12288,0.05047022302945455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3584,1024,0.15080532762739393
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,10240,0.04450666573312548
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,10240,0.22689421971638998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,65536,1.473372459411621
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,10240,0.32618488205803764
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3072,7168,0.06570844517813788
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,8192,0.17871556017133924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,8192,0.03659910957018534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,8192,0.2611688772837321
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3072,6144,0.05534133315086365
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,7168,0.032744000355402626
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,7168,0.15545866224500868
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,7168,0.2332355446285672
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3072,5120,0.04612177941534254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,6144,0.028972443607118394
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,6144,0.13145421610938177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,6144,0.2021182245678372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3072,4096,0.036122666464911565
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,5120,0.025655110677083332
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,5120,0.1737884415520562
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3072,3584,0.03193155593342251
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,4096,0.0907084412044949
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,4096,0.021078222327762183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,12288,0.39123111300998265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3072,3072,0.029338667790095013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,3584,0.08147111203935412
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,3584,0.019634667370054457
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,3584,0.14078844918145075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3072,2560,0.024288000331984624
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,3072,0.14245066377851698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,2560,0.061068442132737905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,2560,0.1430177821053399
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3072,2048,0.020317332612143624
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,2560,0.015127999915017022
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,2048,0.04922577738761902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,65536,1.8455670674641926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3072,1536,0.014896000425020853
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,2048,0.013611555927329592
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,5120,0.110189331902398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,1536,0.039480888181262545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,1536,0.14503910806443956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3072,1024,0.011085333095656501
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,1024,0.029969778325822618
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,1024,0.14789066049787733
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3072,768,0.009242666264375051
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,1024,0.010748444332016839
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,3072,0.07112889157401191
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,768,0.025586666332350835
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3072,512,0.007996444072988298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,768,0.010029333333174387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,768,0.1430435578028361
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,512,0.020921778347757127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,512,0.009637333452701569
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,512,0.1455333365334405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3072,256,0.006917333437336816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,2048,0.14100444316864014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,256,0.008848889006508721
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3072,128,0.005932444499598608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,256,0.14461421966552734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,128,0.01717688971095615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,128,0.009008888569143083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,128,0.1337182256910536
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3072,64,0.005530666559934616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3072,32,0.00590844452381134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,3072,4096,0.14909333652920193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,64,0.01715377800994449
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,32,0.01682222220632765
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,3072,8192,0.07376977470186022
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2560,65536,0.39203111330668133
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,65536,0.2254826757642958
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,16384,0.3588764402601454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,3072,256,0.017885333961910672
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,16384,0.06326399909125434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2560,12288,0.08999466896057129
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,16384,0.4948906898498535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,1536,0.012079111403889127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,12288,0.04462933209207323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2560,10240,0.07040444347593519
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,12288,0.2586497730678982
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,65536,1.3808648851182725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2560,8192,0.05578666925430298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,12288,0.3823208808898926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,10240,0.038916445440716214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,10240,0.21500533156924775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,3072,3072,0.016733333468437195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2560,7168,0.04678044385380215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,8192,0.0322088897228241
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,8192,0.16968888706631133
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,8192,0.2572293281555176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2560,6144,0.03978666663169861
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,7168,0.0292195545302497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,7168,0.23248889711168078
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2560,5120,0.03677599959903293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,6144,0.026310222016440496
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,6144,0.12439378102620442
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,65536,1.837683571709527
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,6144,0.20047199726104736
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2560,4096,0.029088000456492107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,5120,0.02328266700108846
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,5120,0.1732471121682061
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,5120,0.10417511065800984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2560,3584,0.0258559998538759
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,4096,0.08608000146018134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,4096,0.019837333096398246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,4096,0.14914133813646105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2560,3072,0.022788445154825848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,3584,0.076365331808726
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,3584,0.14037511083814833
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,3584,0.018052443861961365
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2560,2560,0.018900444110234577
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,3072,0.01646399994691213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2560,16384,0.1048168871137831
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,3072,0.06435910860697429
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2560,2048,0.016137777103318107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,2560,0.05611111058129204
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,10240,0.32071465916103786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,2560,0.014836443795098199
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,2560,0.14017421669430202
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2560,1536,0.013145777914259167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,7168,0.14634311199188232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,2048,0.04623466730117798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,2048,0.14149867163764104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2560,1024,0.010049777726332346
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,1536,0.03771200113826328
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,1536,0.011748444702890186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,1024,0.0281333327293396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2560,768,0.008647110727098253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,1024,0.010384889112578498
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,1024,0.14446843994988337
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2560,512,0.007267555428875818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,768,0.024259555670950148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,768,0.009866666462686326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,768,0.14705155955420599
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2560,256,0.00628977765639623
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,512,0.009348444640636444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,256,0.017249777913093567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,256,0.1398071050643921
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,256,0.008973333570692275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,3072,0.14797243807050917
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,128,0.016502221425374348
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,128,0.141031993760003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,128,0.008776889079146916
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2560,2048,0.013454222016864352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2560,32,0.005252444495757421
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,1536,0.14990844991472033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,64,0.016166221764352586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,32,0.016162667009565566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2048,65536,0.30810221036275226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2560,512,0.020032889313167997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2560,512,0.14109688334994847
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2048,16384,0.0895306666692098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,65536,0.1806257830725776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,16384,0.04943822158707512
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2560,128,0.005315555466545953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,16384,0.3216799894968669
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2048,12288,0.07542044586605497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2048,10240,0.05039377676116096
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,12288,0.03675555520587497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,12288,0.24988444646199545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2560,64,0.005261333452330695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,12288,0.34914665751987034
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2048,8192,0.0467564430501726
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,10240,0.20169599850972494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,10240,0.03270666797955831
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2048,7168,0.04122666517893473
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,65536,1.2800328996446397
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,8192,0.02626577847533756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,8192,0.24401243527730307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2048,6144,0.03546222382121616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,7168,0.024181332853105333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,7168,0.21875554985470244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,6144,0.11672355069054498
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,6144,0.021967111362351313
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,6144,0.18805600537194145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2048,5120,0.028484443823496502
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,65536,1.6939297782050238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,5120,0.019727110862731934
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,5120,0.09778666496276855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2048,4096,0.02460533380508423
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,16384,0.4570924441019694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,4096,0.016555555992656283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,4096,0.14716445075141057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2048,3584,0.022583999567561682
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,4096,0.08042666647169325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2048,3072,0.01976355579164293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,3584,0.015507555670208402
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,3584,0.07134399811426799
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,3584,0.14560800128512913
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2048,2560,0.017436444759368896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,10240,0.2996239927079943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,3072,0.013855111267831592
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,3072,0.1409671174155341
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,8192,0.15970667203267416
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2048,2048,0.014814222852389017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,2560,0.05267111129230923
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,2560,0.012791111237472959
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,7168,0.138044450018141
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,2048,0.042747557163238525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2048,1536,0.012058667010731168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,2048,0.14067467053731283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,2048,0.01109333336353302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2048,1024,0.009333333207501305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,1536,0.03477511141035292
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,1536,0.010372444159454769
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,1536,0.15023199717203775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2048,768,0.008296888735559251
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,1024,0.026752000053723652
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,1024,0.009445333646403419
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,1024,0.13922311200035944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,5120,0.16275289323594835
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2048,512,0.006903111106819577
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,768,0.022866666316986084
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,768,0.009012444151772393
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,768,0.14719200134277344
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2048,256,0.0052595556610160405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,512,0.008387555678685507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,512,0.018929777873886954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2048,128,0.0052426668504873914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,256,0.016542222764756944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,256,0.00830577810605367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,128,0.01556799974706438
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,128,0.13924445046318903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2048,64,0.004547555413511065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,2048,32,0.004861333303981357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,2560,0.14613600571950278
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,32,0.015228443675571017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1536,65536,0.294269323348999
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,512,0.13995644781324598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,2048,256,0.1412986649407281
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,3072,0.06251733170615302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1536,16384,0.07271289163165622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,2048,128,0.008037333687146505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,2048,64,0.015498666299713982
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,16384,0.04557333389918009
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1536,12288,0.059226665231916636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,65536,1.2009092966715496
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,16384,0.30704977777269155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1536,10240,0.04770400126775106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,12288,0.03545599844720628
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,16384,0.45305776596069336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,12288,0.23138756222195098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1536,8192,0.03766577773623996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,12288,0.3483697838253445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,10240,0.031191110610961914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,10240,0.193503114912245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1536,7168,0.03379111157523261
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,10240,0.29652267032199436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,8192,0.025659556190172832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,8192,0.15277689033084446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1536,6144,0.029567112525304157
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,7168,0.023680888944202002
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,7168,0.13083822197384304
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,7168,0.21384621991051567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,6144,0.11120177639855279
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,65536,1.6901031070285375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,6144,0.1857386695014106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,6144,0.021292444732454088
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1536,4096,0.020304888486862183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,5120,0.09324177768495347
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,65536,0.16990666919284395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,5120,0.019000889526473153
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,5120,0.16075555483500162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1536,3584,0.018585777944988675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,4096,0.07607733541064791
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,4096,0.01623644431432088
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1536,3072,0.015997333659066092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,3584,0.0151555554734336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,3584,0.14693155553605822
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1536,2560,0.014196443888876172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,3072,0.05873066849178738
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,3072,0.14368800322214761
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,3072,0.013782222237851886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1536,2048,0.01238844460911221
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,2560,0.049399998452928334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,2560,0.012431999875439538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,2560,0.14640266365475124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1536,1536,0.009317333499590555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,2048,0.04141511188613044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,2048,0.011044444309340583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,2048,0.13958311080932617
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1536,1024,0.008271111382378472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,8192,0.24018756548563638
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,1536,0.03344355689154731
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,1536,0.010332444475756751
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,1536,0.146814227104187
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1536,768,0.006774222271309958
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,1024,0.025723555021815833
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,1024,0.009357333183288574
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1536,5120,0.02405422263675266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,1024,0.1408604383468628
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,768,0.021958221991856892
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1536,512,0.005943999936183293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,768,0.14356177382998997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,768,0.008974221845467886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,512,0.008364444805516137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,512,0.1396426624721951
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,256,0.015494222442309061
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,256,0.1403599977493286
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,4096,0.148561782307095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,256,0.008204444415039485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,128,0.01479555500878228
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1536,64,0.004176889028814104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1536,128,0.007993777592976889
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1536,32,0.004176889028814104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,64,0.014448889427714877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,32,0.014236445228258768
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1024,65536,0.20023555225796172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1536,256,0.004860444615284602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,512,0.017901332841979135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1024,16384,0.048842665221956044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1536,128,0.00423555572827657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,65536,0.12326755788591172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1536,3584,0.06758222315046522
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1024,12288,0.039232889811197914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1536,128,0.14170043998294407
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,16384,0.28626754548814565
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,65536,1.1033040152655709
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1024,10240,0.030982222821977403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,16384,0.42214756541781956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,12288,0.2161795563167996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,12288,0.026714666022194758
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1024,8192,0.02460533380508423
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,12288,0.3265048927730984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,10240,0.0237128883600235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1024,7168,0.022247110803922016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,10240,0.18040888839297822
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,8192,0.019469334019554984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,8192,0.14197422398461237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,8192,0.2220391167534722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1024,6144,0.019893333315849304
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,7168,0.017851556340853374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,7168,0.12280978096856011
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,7168,0.19975645012325713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,65536,1.5509973102145727
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1024,5120,0.016733333468437195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,6144,0.10340266757541233
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,6144,0.17136177751753065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1024,4096,0.014325333966149224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,5120,0.01517688896920946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,5120,0.08562933074103461
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1024,3584,0.012940444052219391
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,4096,0.14424444569481745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,4096,0.07028888993793063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,16384,0.03424177898301019
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,3584,0.062146663665771484
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1024,3072,0.011424889167149862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,3584,0.012098666694429187
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,3584,0.14777244461907282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1024,2560,0.01036088913679123
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,3072,0.011023999916182624
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,3072,0.05336622065967984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,3072,0.14533066749572754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1024,2048,0.008941333327028487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,2560,0.04604622059398227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,2560,0.010397333237859938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,2560,0.14390133486853704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1024,1536,0.0075902218619982404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,2048,0.03875999980502658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,10240,0.27223022778828937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,2048,0.1469351053237915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1024,1024,0.006145777801672618
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,1536,0.03054311209254795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,1536,0.009015999734401703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,1024,0.023997333314683702
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1024,768,0.005416888743638992
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,6144,0.016534222496880423
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,5120,0.15054133203294542
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,768,0.020610666937298246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,768,0.14643466472625732
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,4096,0.013064888616402945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1024,512,0.00463377767139011
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,768,0.00796088907453749
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1024,256,0.0042364444169733255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,512,0.016878222425778706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,512,0.007652444144090016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,512,0.14213511678907606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1024,128,0.0038773334688610504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,256,0.014472888575659858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,256,0.007298666569921706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,256,0.14547377162509495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1024,64,0.003677333394686381
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,128,0.013738666971524557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,128,0.13292177518208823
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,1024,32,0.003530666646030214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,128,0.007299555672539606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,64,0.01348622226052814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,2048,0.009650666680600908
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,768,65536,0.16101245085398355
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,768,16384,0.04458399944835239
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,1536,0.1467075612809923
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,65536,0.12177688545650905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,1024,1024,0.00832799987660514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,768,12288,0.03156266609827677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,16384,0.03403288788265652
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,768,16384,0.2788453367021349
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,16384,0.412733342912462
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,768,12288,0.20876977178785536
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,12288,0.026548445224761963
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,768,10240,0.023711999257405598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,12288,0.32502932018703884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,768,65536,1.0655724207560222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,10240,0.022944889134830896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,768,8192,0.020818667279349435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,768,10240,0.17399288548363578
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,10240,0.27047732141282826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,768,7168,0.01791555517249637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,1024,32,0.013471999930010902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,768,8192,0.1364284488889906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,8192,0.22293421957227919
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,768,6144,0.01650311052799225
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,7168,0.01792355544037289
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,768,7168,0.11810844474368626
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,7168,0.2002506653467814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,65536,1.5383937623765733
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,1024,1024,0.1436986658308241
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,6144,0.016492444607946608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,6144,0.17143555482228598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,768,4096,0.011852444046073489
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,5120,0.1516231165991889
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,5120,0.01519377695189582
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,768,4096,0.06818577978346083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,4096,0.013096888860066732
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,4096,0.14371644126044378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,768,3584,0.01088355564408832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,3584,0.012151999606026543
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,3584,0.14272533522711858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,768,3072,0.009705777797434065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,3072,0.14036888546413845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,8192,0.0192577772670322
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,768,2560,0.008842666943868002
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,3072,0.011250666446155973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,768,2560,0.04309511184692383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,2560,0.1405857801437378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,768,5120,0.014114666316244336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,768,2048,0.007989333735571967
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,768,6144,0.09939110941357082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,768,1536,0.0069448889957533936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,2560,0.010671111444632212
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,768,2048,0.03577155537075467
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,2048,0.1404577758577135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,768,5120,0.08277955320146349
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,768,1024,0.005936000082227919
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,768,1536,0.02938844429122077
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,1536,0.009034666750166152
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,1536,0.14372089174058703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,768,768,0.0052728888889153796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,768,1024,0.023005333211686876
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,1024,0.14196799861060247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,1024,0.008309333688682979
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,768,512,0.004519999855094486
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,768,768,0.01995111174053616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,768,0.00794311116139094
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,768,0.14059466785854763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,768,3584,0.059782220257653125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,768,256,0.004518222063779831
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,768,512,0.016551999582184684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,512,0.14025866985321045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,768,128,0.004222222086456087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,768,256,0.014139556222491793
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,256,0.007263110743628607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,256,0.1449297799004449
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,768,64,0.0038524443904558816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,768,128,0.013388444152143268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,768,32,0.004006222304370669
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,128,0.007290666302045186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,768,128,0.1420231130388048
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,768,64,0.013095999757448832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,768,32,0.012899555265903473
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,512,16384,0.03492000036769443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,65536,0.12068888876173232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,2048,0.009693333672152625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,512,16384,0.2656071186065674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,512,12288,0.024490666058328416
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,16384,0.033471110794279314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,16384,0.4077591101328532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,512,12288,0.20075378153059217
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,12288,0.02588444451491038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,512,10240,0.020419556233617995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,12288,0.31861066818237305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,512,65536,1.0237840016682942
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,768,3072,0.0499199993080563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,10240,0.022663111488024395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,512,10240,0.16779111491309273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,10240,0.2707742320166694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,512,65536,0.11328978008694118
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,512,8192,0.13135466310713026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,8192,0.018840889135996502
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,8192,0.21787110964457193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,512,6144,0.013723555538389417
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,7168,0.01755555636352963
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,7168,0.19995289378696016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,512,6144,0.09291644228829278
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,6144,0.015981333123313058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,6144,0.16930754979451498
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,5120,0.14856799443562826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,768,512,0.007606222397751278
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,512,8192,0.01777866648303138
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,5120,0.014767110347747803
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,512,4096,0.010384889112578498
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,512,7168,0.01573511130279965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,4096,0.14006488853030735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,512,4096,0.06588088803821139
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,512,3584,0.009761778016885122
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,4096,0.012621333201726278
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,512,3584,0.05727377865049574
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,512,7168,0.11281066470675999
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,3584,0.14184533225165472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,512,3072,0.008629333641793992
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,3584,0.011759999725553723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,512,2560,0.0079644446571668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,512,3072,0.04894489049911499
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,3072,0.14044533835517034
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,512,5120,0.012108444339699216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,512,2048,0.006986666884687211
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,512,2560,0.04186044467820061
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,2560,0.010135999984211391
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,2560,0.14911377429962158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,512,5120,0.0797262191772461
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,512,1536,0.006224000205596288
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,512,2048,0.034103999535242714
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,2048,0.009391999906963771
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,2048,0.14080888695187038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,512,1024,0.0052426668504873914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,512,1536,0.028415110376146104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,1536,0.14760533968607584
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,512,768,0.004554666578769684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,512,1024,0.022413333257039387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,1024,0.1409297784169515
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,1024,0.008011555506123437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,512,512,0.0041751112374994485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,768,0.007648888561460707
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,512,768,0.01956888867749108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,768,0.14756177531348333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,512,256,0.00351200004418691
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,512,512,0.016169778174824186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,512,0.0074337778819931885
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,512,128,0.0031866667171319327
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,512,256,0.013741333451535968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,65536,1.524780485365126
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,256,0.14603288968404135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,256,0.007259555160999298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,512,64,0.0032097777972618737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,512,32,0.0032097777972618737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,512,128,0.013112000293201871
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,128,0.1414417823155721
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,128,0.007317333585686154
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,512,64,0.0127688886390792
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,256,65536,0.0653031137254503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,3072,0.010791111323568555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,512,32,0.012423111332787408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,256,16384,0.019024888674418133
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,65536,0.120432005988227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,16384,0.03229599859979417
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,256,12288,0.018361777067184448
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,256,16384,0.2540568775600857
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,512,1536,0.00868622213602066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,256,12288,0.19258311059739855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,256,10240,0.016228444046444364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,12288,0.3212249014112684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,256,10240,0.16045422024197048
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,10240,0.022005332840813532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,512,512,0.14160354932149252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,256,8192,0.013493333425786761
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,256,65536,0.9789288838704427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,256,7168,0.012358222570684222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,8192,0.018566222654448617
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,256,8192,0.125946667459276
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,8192,0.21507110860612658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,256,6144,0.01111822244193819
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,256,7168,0.10754222340053982
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,7168,0.017403556240929496
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,7168,0.19684444533454049
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,256,5120,0.009970666633711921
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,65536,1.5166195763481989
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,6144,0.0159262220064799
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,6144,0.16929866207970512
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,256,4096,0.00832622249921163
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,256,5120,0.07600533299975924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,5120,0.014832889040311178
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,5120,0.14956088860829672
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,16384,0.40789421399434406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,256,3584,0.008088889221350351
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,12288,0.024719999896155462
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,4096,0.012631110846996307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,4096,0.14135555426279703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,256,3072,0.007262222468852997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,3584,0.011722666521867117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,256,3072,0.04629333482848274
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,256,2560,0.006618666566080517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,10240,0.27105598979526097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,3072,0.010964444114102257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,2560,0.010062221851613786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,256,2560,0.038557334078682795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,256,2048,0.005588444570700328
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,256,2048,0.03255555695957608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,2048,0.1435742245780097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,256,6144,0.0908400019009908
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,256,1536,0.005222222043408289
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,256,1536,0.027220444546805486
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,256,4096,0.06200622187720405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,1536,0.14114844799041748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,1536,0.00867377801073922
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,3584,0.1501120063993666
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,256,1024,0.00452533322903845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,3072,0.14367732736799452
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,256,1024,0.021612443857722815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,256,768,0.0038444445365005066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,1024,0.15148533715142146
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,768,0.007786666353543599
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,256,768,0.01871466636657715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,768,0.14557422531975642
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,256,512,0.015536889433860779
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,512,0.1440417766571045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,256,256,0.0032133333798911837
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,512,0.007419555551475949
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,2048,0.009307555854320526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,256,256,0.013087111214796701
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,256,0.007253333098358578
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,256,128,0.003156444471743372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,256,0.1463466617796156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,128,0.0070186663005087115
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,256,128,0.012431110772821637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,128,0.13697155316670737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,256,64,0.0028284444577164124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,256,32,0.0028195555011431375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,256,3584,0.0544453329510159
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,128,65536,0.05356711149215698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,256,64,0.01236177815331353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,256,32,0.012087999946541257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,256,1024,0.00797333319981893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,128,16384,0.013111111190583972
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,65536,0.12041600545247395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,256,2560,0.1447733375761244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,256,512,0.0038888889054457345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,128,12288,0.013735999663670858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,128,16384,0.25405777825249565
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,16384,0.03231822119818793
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,128,10240,0.012713777522246042
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,12288,0.024772443705134924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,16384,0.41307732794019913
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,128,8192,0.013014222184816996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,128,12288,0.19232000244988334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,12288,0.3203226725260417
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,10240,0.021917333205540974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,128,10240,0.1605608860651652
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,128,7168,0.011735999749766456
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,10240,0.2699208789401584
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,8192,0.01850844422976176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,128,8192,0.125617782274882
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,8192,0.2140337758594089
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,128,6144,0.0107022225856781
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,128,65536,0.9783555136786567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,128,7168,0.10758222473992242
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,7168,0.017279111676745944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,128,5120,0.009686222506894005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,7168,0.19581955009036592
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,128,4096,0.007986666427718269
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,6144,0.16887733671400282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,5120,0.014840889308187695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,5120,0.14621688259972468
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,128,5120,0.07575555642445882
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,128,3584,0.007628444168302748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,128,4096,0.06211733155780368
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,4096,0.14524088965521917
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,4096,0.012764444781674279
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,128,3072,0.006884444504976273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,65536,1.512472046746148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,128,3584,0.05365777677959866
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,3584,0.13938755459255642
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,3584,0.011690666278203329
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,128,2560,0.006200888918505774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,128,3072,0.0462719996770223
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,3072,0.010805333654085795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,3072,0.14556533760494655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,128,2048,0.005403555515739653
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,128,2560,0.03927466604444716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,2560,0.010029333333174387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,2560,0.14119822449154323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,2048,0.009342222577995723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,2048,0.14542578326331243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,128,1024,0.0041804446114434134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,1536,0.14080888695187038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,1536,0.00867911097076204
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,128,1024,0.02129511038462321
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,128,768,0.003859555555714501
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,128,768,0.018589332699775696
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,6144,0.015840889679061044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,128,6144,0.09035378032260472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,128,512,0.003524444583389494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,768,0.007638221813572778
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,768,0.14495821793874106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,128,512,0.015170666906568738
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,128,256,0.0032151111712058387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,512,0.007324444750944774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,512,0.15081600348154703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,128,256,0.01276622215906779
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,256,0.14537599351671007
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,128,128,0.003167111012670729
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,256,0.007248889240953658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,128,2048,0.03221866819593642
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,128,1536,0.0270471109284295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,128,128,0.012099555797047086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,128,64,0.002854222224818336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,128,0.14457510577307806
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,128,0.00720977783203125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,128,32,0.0028444443725877335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,64,65536,0.04843822121620178
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,128,64,0.012076444096035428
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2048,128,1024,0.008205333517657386
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,128,32,0.01180622229973475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,64,16384,0.015238222148683337
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,64,10240,0.011420444481902652
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,64,16384,0.25371821721394855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,64,12288,0.19197600417666963
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,64,10240,0.15973599751790366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,64,8192,0.010057777994208867
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,64,7168,0.009321778184837764
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,64,6144,0.009020444419648912
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,64,8192,0.1253768867916531
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,64,7168,0.1081297794977824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,64,5120,0.008262222011884054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,64,6144,0.09011555380291408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,64,4096,0.007633777956167857
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,64,3584,0.0068986668354935115
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,128,1536,0.004877333425813251
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,64,65536,0.9776977962917752
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,64,5120,0.07572711176342435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,64,3072,0.006559111177921295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,64,4096,0.061350219779544406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,64,2560,0.005864000154866113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,64,3584,0.05334844523006015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,64,1536,0.004552000098758274
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,64,2560,0.03916622201601664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,64,3072,0.046217779318491616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,64,2048,0.03222489025857713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,64,1024,0.004206222378545337
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,64,768,0.003842666745185852
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,64,1536,0.02674666709370083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,64,512,0.0034933334423436057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,2048,128,1024,0.14113510979546443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,64,256,0.0031999999450312722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,64,1024,0.021306667062971327
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,64,128,0.0028106667515304354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,64,12288,0.012806221842765808
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,64,768,0.018581334087583754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,64,64,0.002797333316670524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,64,32,0.0028613333900769553
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,64,512,0.015160888433456421
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,64,256,0.012804444465372296
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,32,65536,0.04717777834998237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,64,32,0.011752888560295105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,32,12288,0.008657777474986183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,32,16384,0.010023111270533668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,32,10240,0.008170666793982187
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,32,8192,0.012084444363911947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,32,12288,0.19194222821129692
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,32,10240,0.1594417757458157
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,32,7168,0.01108266661564509
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,32,8192,0.12514044178856745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,32,6144,0.010392888552612728
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,32,5120,0.009360888765917884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,32,7168,0.10717244280709161
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,32,4096,0.008623110751310984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,32,6144,0.09054044220182632
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,64,2048,0.005225777626037598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,32,3584,0.0079297779334916
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,32,5120,0.0755680004755656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,32,3072,0.006540444576077991
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,32,2560,0.00590844452381134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,32,4096,0.06025333536995781
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,32,3584,0.053242666853798755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,32,2048,0.005250666704442766
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,32,3072,0.045893334680133395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,32,1536,0.004567111117972268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,32,2560,0.038944890101750694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,32,2048,0.03219644559754266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,32,1536,0.026937777797381084
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,32,768,0.0035111109415690103
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,32,1024,0.021301334102948506
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,64,128,0.012359110845459832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,32,512,0.003522666792074839
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,64,64,0.012060444388124677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,32,256,0.003132444495956103
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,32,128,0.0028373334142896864
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,32,768,0.018583110637134977
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,32,64,0.0028231110837724474
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,32,512,0.015375110838148328
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,32,32,0.0028124445428450904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,32,256,0.012792000340090858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,32,16384,0.2532835536532932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,32,128,0.012086222569147745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,32,64,0.012089778151777057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,32,32,0.011694221860832639
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2048,32,65536,0.9766808615790473
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2048,32,1024,0.0038862224254343244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,16384,0.43777865833706325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,65536,12288,0.8786062134636773
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,65536,16384,1.2483057445949979
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,12288,1.0504560470581055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,16384,1.5120844311184352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,12288,0.32642579078674316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,65536,10240,0.7617404195997449
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,10240,0.28388089603847927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,10240,0.917927106221517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,65536,8192,0.6610328886244032
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,8192,0.7122542063395182
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,12288,1.0528230667114258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,8192,0.2267679903242323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,65536,7168,0.5540942086113824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,16384,1.4102747175428603
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,10240,0.8078515794542102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,7168,0.6093769073486328
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,65536,6144,0.4718551105923123
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,7168,0.20146577888064912
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,6144,0.1730417807896932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,8192,0.6783191363016764
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,6144,0.5740310880872939
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,5120,0.14860888322194418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,5120,0.4943644205729167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,65536,4096,0.3426400025685628
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,7168,0.6051884227328831
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,6144,0.5160951084560818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,4096,0.36005155245463055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,4096,0.124908447265625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,5120,0.4362471103668213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,65536,3584,0.3115359942118327
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,3584,0.3250853220621745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,3584,0.11349066098531087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,65536,3072,0.23817245165506998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,4096,0.3548204369015164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,3072,0.26828622817993164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,65536,5120,0.35735199186537003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,65536,2560,0.18608621756235758
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,3072,0.09871644443935818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,3584,0.3264097902509901
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,2560,0.23867556783888075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,65536,2048,0.1566151115629408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,2560,0.08774844143125747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,65536,1536,0.12483999464246963
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,2048,0.19235377841525605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,3072,0.4045768843756781
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,2048,0.07353244225184123
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,2048,0.2148053381178114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,65536,1024,0.08048622475730048
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,1536,0.15092621909247503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,1536,0.05823644664552477
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,1536,0.18209866682688394
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,65536,768,0.07708622349633111
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,1024,0.046557333734300405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,1024,0.11705866124894883
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,65536,512,0.04807733496030172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,1024,0.15217778417799208
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,768,0.09346399704615276
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,768,0.14685600333743626
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,65536,256,0.03815733393033346
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,512,0.08125066757202148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,512,0.038006222910351224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,65536,128,0.03149688906139798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,256,0.1465324428346422
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,128,0.03412977854410807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,128,0.06911910904778375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,128,0.13830133279164633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,65536,64,0.02678311036692725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,65536,32,0.02738222148683336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,64,0.06878133614857991
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,32,0.06875199741787381
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,2560,0.24928532706366646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,768,0.04170577724774679
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,65536,512,0.14455733034345838
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,65536,256,0.034879999028311834
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,16384,16384,0.3178035683102078
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,16384,65536,1.2659804026285808
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,65536,0.47928264405992294
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,16384,0.12029777632819282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,16384,0.4375528759426541
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,16384,12288,0.26284888055589467
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,12288,0.09286755323410034
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,16384,10240,0.2127671109305488
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,16384,0.4679270850287543
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,12288,0.3458213276333279
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,65536,1.8186168670654297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,16384,8192,0.16083289517296687
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,10240,0.07877600193023682
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,10240,0.26766755845811635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,65536,256,0.07372088564766778
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,16384,7168,0.16506755352020264
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,8192,0.21342399385240343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,8192,0.06398489077885945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,10240,0.3156239986419678
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,16384,6144,0.14633954895867243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,8192,0.25827733675638836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,7168,0.18997332784864637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,16384,5120,0.10437599817911784
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,7168,0.23724177148607042
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,6144,0.049357334772745766
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,6144,0.16606577237447104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,6144,0.20327644877963594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,5120,0.14092889097001818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,16384,4096,0.09601155254575942
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,5120,0.18250932958390978
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,4096,0.1152977810965644
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,4096,0.1540586683485243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,12288,0.3581857681274414
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,16384,3584,0.06876266664928861
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,65536,1.8823074764675563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,3584,0.10233333375718857
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,3584,0.031094223260879517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,3584,0.1497911082373725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,16384,3072,0.0698888897895813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,16384,2560,0.05766666597790188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,7168,0.05770577987035116
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,3072,0.14449066585964626
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,2560,0.026233777403831482
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,2560,0.14523200194040933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,2560,0.08089066876305474
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,16384,2048,0.04222933451334635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,5120,0.043248001072141856
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,2048,0.1448542277018229
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,2048,0.06564533048205905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,2048,0.021943999661339655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,4096,0.035924444595972695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,1536,0.01904177831278907
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,1536,0.0539973345067766
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,1536,0.1445831060409546
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,16384,1024,0.025791111919615004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,16384,768,0.019834667444229126
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,1024,0.041518221298853554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,1024,0.015895111693276297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,1024,0.1458266708585951
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,16384,512,0.015240889456537036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,768,0.03536622060669793
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,768,0.014119111829333834
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,768,0.1422684457567003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,16384,256,0.012638222012254926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,512,0.02868000004026625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,3072,0.09072266684638129
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,512,0.013035555680592855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,512,0.14540444480048284
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,16384,128,0.01037599974208408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,256,0.02602400051222907
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,256,0.012412444584899478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,256,0.14493600527445474
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,16384,64,0.008994667066468133
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,128,0.02513066596455044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,16384,32,0.009317333499590555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,128,0.0120666664507654
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,64,0.025072889195548162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,16384,32,0.024683554967244465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,16384,1536,0.03919733232922024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,12288,16384,0.2800399992201063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,16384,0.35320266087849933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,65536,0.3769804371727838
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,12288,65536,0.9759662416246203
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,16384,3072,0.02795733345879449
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,12288,12288,0.21609955363803438
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,16384,0.429749329884847
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,16384,128,0.1372213363647461
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,12288,0.08062044117185804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,12288,0.28104710578918457
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,12288,0.32394221093919545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,65536,1.4918382432725694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,10240,0.21584888299306235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,10240,0.06252266963322957
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,12288,8192,0.1356017722023858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,8192,0.05133777856826782
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,10240,0.2907902134789361
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,12288,7168,0.11642489168379043
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,8192,0.23039555549621582
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,7168,0.04584533307287428
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,12288,6144,0.09993955824110244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,7168,0.15668533907996282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,7168,0.2103591097725762
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,16384,0.0952684415711297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,12288,5120,0.0860933330323961
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,6144,0.14198933707343206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,6144,0.18762577904595268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,12288,4096,0.07117333014806111
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,5120,0.03583555420239767
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,5120,0.1156160036722819
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,5120,0.1628533336851332
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,12288,10240,0.17594310972425672
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,12288,3584,0.06118577718734741
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,4096,0.09564089112811619
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,4096,0.029218667083316382
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,65536,1.6236862606472438
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,4096,0.14820444583892822
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,3584,0.08562399943669637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,12288,3072,0.05853333075841268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,3584,0.14619110690222845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,3584,0.02664711078008016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,12288,2560,0.04710400104522705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,3072,0.14816444449954563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,3072,0.02405422263675266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,3072,0.07702489031685723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,8192,0.17605777581532797
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,12288,2048,0.03635911146799723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,2560,0.06586133109198676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,2560,0.021689777572949726
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,2560,0.14640711413489446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,12288,1536,0.028224888775083754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,2048,0.1402382188373142
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,2048,0.018746667438083224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,12288,1024,0.019951999187469482
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,1536,0.0443102220694224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,1536,0.01625777781009674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,1536,0.1455813381406996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,12288,768,0.016198222835858662
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,1024,0.0342497792508867
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,1024,0.14957777659098306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,6144,0.04043466515011258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,12288,512,0.012686221963829465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,768,0.029419554604424372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,768,0.012402666939629449
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,768,0.14611732959747314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,12288,256,0.011983111500740051
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,512,0.024736000431908503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,512,0.011630222201347351
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,512,0.14184978273179796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,12288,128,0.010377777947319878
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,256,0.022338666849666174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,256,0.010695999695195092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,12288,64,0.009689778089523315
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,128,0.021700445148679946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,128,0.011067555182509951
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,128,0.13523377312554255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,12288,32,0.00758755538198683
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,64,0.021409778131379023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,32,0.021340444684028625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,12288,2048,0.054436445236206055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,12288,1024,0.013776000175211163
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,65536,0.32773955663045246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,10240,65536,0.8108328713311089
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,12288,256,0.14690221680535212
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,16384,0.08383288648393418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,16384,0.3137297895219591
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,65536,1.2976648542616103
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,12288,0.06250311268700494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,16384,0.4109066592322455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,12288,0.2362968921661377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,10240,10240,0.15965955787234837
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,12288,0.3070053259531657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,10240,0.054515557156668774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,10240,8192,0.13319378428988987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,10240,0.26002756754557294
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,8192,0.045646223757002086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,10240,16384,0.25137421819898814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,8192,0.2142124440934923
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,10240,7168,0.11590933799743652
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,10240,6144,0.10159466663996379
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,7168,0.14161511262257895
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,7168,0.040067556831571795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,7168,0.1927493280834622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,10240,5120,0.0850293305185106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,6144,0.12853333685133192
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,65536,1.47399287753635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,6144,0.17547644509209526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,10240,12288,0.18086666531032988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,10240,4096,0.06413866413964166
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,5120,0.10406844483481513
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,5120,0.15635466575622559
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,5120,0.031352887550989784
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,4096,0.08695377906163533
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,10240,3584,0.05579466952217949
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,4096,0.141412443584866
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,10240,3072,0.047394666406843394
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,10240,0.19422933790418836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,3584,0.0776968863275316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,3584,0.02344888945420583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,3584,0.14890666802724203
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,10240,2560,0.041689776711993747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,3072,0.021439111895031396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,3072,0.06821778085496691
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,3072,0.14162311289045545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,10240,2048,0.03138400117556254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,2560,0.14671644899580213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,2560,0.019341333044899833
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,2048,0.049068444304996066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,10240,1536,0.025690666503376428
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,2048,0.14085777600606283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,1536,0.04042844308747186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,1536,0.01423377792040507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,1536,0.14986933602227107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,10240,1024,0.017321777012613084
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,6144,0.035511109564039446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,10240,768,0.014456889695591398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,1024,0.0122079998254776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,1024,0.03119200136926439
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,1024,0.150017778078715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,10240,512,0.01148444414138794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,768,0.02680444386270311
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,768,0.01127822200457255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,768,0.1403706603580051
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,10240,256,0.008962666822804345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,4096,0.025800888737042744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,512,0.022650667362742957
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,512,0.1467404497994317
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,256,0.020629333125220407
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,256,0.14497955640157065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,10240,128,0.007385777930418651
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,8192,0.1558026605182224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,2560,0.05942044655481974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,128,0.020277332928445604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,10240,128,0.13393599457210964
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,10240,64,0.008637333081828224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,10240,32,0.008599999878141616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,2048,0.016963554753197562
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,32,0.0196142229768965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,10240,64,0.019934222102165222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,8192,65536,0.5766853226555718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,65536,0.2679564423031277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,8192,16384,0.16362667083740234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,512,0.010687111152542962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,256,0.010039110978444418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,16384,0.31577245394388836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,16384,0.06661688619189791
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,8192,12288,0.1368488868077596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,10240,128,0.009992000129487779
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,12288,0.051559110482533775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,65536,1.1433582305908203
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,8192,10240,0.1120835542678833
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,12288,0.20210844940609404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,16384,0.3613288932376438
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,8192,8192,0.07804622252782185
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,12288,0.281384891933865
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,10240,0.16823911666870117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,8192,0.1370613310072157
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,8192,0.03686755564477708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,10240,0.2303102281358507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,8192,0.19551200336880156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,8192,6144,0.06323555443021986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,7168,0.03323910964859857
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,7168,0.12073155244191487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,7168,0.17697066730923125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,8192,5120,0.05019911130269369
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,6144,0.10681333144505818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,6144,0.029699554045995075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,6144,0.15910577774047852
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,8192,4096,0.0502106679810418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,5120,0.09113155470954047
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,5120,0.02637600070900387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,5120,0.14623022079467773
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,65536,1.3174151314629448
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,8192,3584,0.043066667185889355
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,4096,0.07414044274224176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,4096,0.14697333176930746
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,4096,0.022249778111775715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,8192,3072,0.03569599986076355
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,3584,0.1491804387834337
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,3584,0.06736355357699923
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,3584,0.020658666888872784
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,3072,0.018585777944988675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,8192,2560,0.03047377864519755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,3072,0.05992977486716377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,8192,2048,0.025215110845035974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,2560,0.05158844590187073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,2560,0.13983200656043157
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,2560,0.016872000363137987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,10240,0.04500533474816216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,2048,0.04281244344181485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,2048,0.1472408903969659
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,2048,0.014874666929244995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,8192,7168,0.08157866530948214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,1536,0.03602044449912177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,1536,0.14584267139434814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,8192,768,0.012469333079126147
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,1024,0.011056000159846412
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,1024,0.14538221889071995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,768,0.02400622268517812
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,768,0.01034222212102678
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,768,0.14520355065663657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,512,0.020983111527231004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,512,0.13906755712297228
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,8192,256,0.007291555404663086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,512,0.00960533320903778
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,256,0.018912000788582694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,3072,0.15248444345262316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,8192,128,0.007619555625650618
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,8192,1536,0.01980444457795885
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,128,0.01854044364558326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,128,0.13277333312564427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,8192,64,0.007269333634111617
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,1024,0.027619555592536926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,128,0.00905866672595342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,8192,32,0.007262222468852997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,64,0.018236445056067575
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,8192,32,0.017867555220921833
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,8192,512,0.010009778042634329
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,7168,65536,0.5762675603230795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,65536,0.24724443753560385
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,8192,256,0.14542667071024576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,7168,16384,0.15767378277248806
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,256,0.009024889104896123
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,7168,12288,0.11718400319417317
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,8192,1024,0.014811555544535318
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,16384,0.06282755401399401
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,8192,1536,0.012860444684823355
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,65536,1.0701333151923287
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,16384,0.3496399985419379
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,12288,0.18802577919430205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,12288,0.04854399959246317
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,7168,10240,0.10717511177062988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,12288,0.26885512140062123
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,7168,8192,0.08003999789555867
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,10240,0.04157600137922499
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,10240,0.15877155462900797
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,10240,0.2281928857167562
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,7168,7168,0.07860266500049166
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,8192,0.1284666723675198
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,8192,0.03473155697186788
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,8192,0.19287644492255318
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,7168,6144,0.07025333245595296
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,7168,0.031150221824645996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,7168,0.1147128873401218
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,7168,0.1749217775132921
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,7168,5120,0.05250933435228136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,6144,0.10110044479370117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,6144,0.028239998552534316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,6144,0.15838044219546848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,65536,1.310174200269911
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,7168,4096,0.04786755641301473
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,5120,0.0865937802526686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,5120,0.14233688513437906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,5120,0.0251893334918552
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,7168,3584,0.03968977928161621
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,4096,0.15162132845984563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,4096,0.021545777718226116
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,3584,0.019724445210562814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,7168,3072,0.03475911087459988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,3584,0.06224266688028971
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,7168,2560,0.02921600143114726
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,3072,0.05577066871854994
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,3072,0.018217777212460835
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,3072,0.14501333236694336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,7168,2048,0.025075554847717285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,2560,0.04871733321083916
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,2560,0.14695200655195448
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,2560,0.016502221425374348
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,7168,1536,0.02047911120785607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,2048,0.014141332772043018
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,2048,0.14093600379096136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,1536,0.03431199987729391
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,7168,1024,0.015470221638679504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,1536,0.012637333737479316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,1536,0.14821155865987143
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,1024,0.026128000683254663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,1024,0.01074222226937612
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,1024,0.14139466815524632
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,7168,768,0.013003555436929068
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,16384,0.2577582200368245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,768,0.010026666853162978
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,768,0.022954665952258643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,7168,512,0.009848000274764167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,4096,0.07018666797214083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,7168,256,0.0075297777851422625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,3584,0.14067999521891275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,512,0.009487999810112847
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,512,0.14721333980560303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,256,0.018180444836616516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,256,0.009007111191749573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,256,0.14140000608232287
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,128,0.017665778597195942
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,7168,64,0.007251555720965068
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,128,0.13667821884155273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,7168,128,0.009003555609120263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,7168,32,0.010159999959998662
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,64,0.01750577820671929
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,2048,0.04095022214783563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,32,0.017268444101015728
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,6144,16384,0.13724533716837564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,6144,65536,0.4611804220411513
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,65536,0.23575377464294434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,16384,0.2366337776184082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,7168,768,0.14043022526635066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,16384,0.057894223266177706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,6144,12288,0.09848444329367743
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,7168,512,0.01960088809331258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,16384,0.3216702143351237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,7168,128,0.006559111177921295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,6144,10240,0.09122666385438706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,12288,0.1713191138373481
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,12288,0.04370489054256015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,6144,8192,0.0682506693734063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,12288,0.2445653279622396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,10240,0.03698222173584832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,10240,0.14710488584306505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,6144,7168,0.06916622320810954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,8192,0.12012622091505264
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,8192,0.032672888702816434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,8192,0.17595467302534315
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,6144,6144,0.05882755253050062
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,7168,0.028852442900339764
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,7168,0.10716711150275336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,7168,0.1609146727455987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,6144,5120,0.04768622252676222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,6144,0.09358755747477214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,6144,0.14621244536505804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,6144,0.02588444451491038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,6144,4096,0.039488001002205744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,5120,0.08091200060314603
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,65536,1.1666409174601238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,5120,0.14732444286346436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,5120,0.023296000228987798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,6144,3584,0.033266666862699724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,4096,0.06475377745098539
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,4096,0.15118666489919028
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,4096,0.02000711030430264
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,3584,0.05761688947677612
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,6144,3072,0.02929688824547662
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,3584,0.14705955982208252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,3584,0.017543110582563613
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,6144,2560,0.024588444166713293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,3072,0.05099822084108988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,65536,0.9784746170043945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,3072,0.14524000220828587
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,6144,2048,0.020993777447276648
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,2560,0.04553066690762838
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,2560,0.014732443624072604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,6144,1536,0.015635554989178974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,2048,0.03820177912712097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,2048,0.14189778433905706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,2048,0.013220444321632385
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,6144,1024,0.012416889270146688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,1536,0.03183022141456604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,1536,0.011123555401961008
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,1536,0.1439173354042901
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,6144,768,0.01071200023094813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,1024,0.024713777833514746
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,1024,0.009649777577983009
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,1024,0.14864355987972683
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,6144,512,0.009143999881214565
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,768,0.021900445222854614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,768,0.009319999979601966
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,512,0.0188755558596717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,512,0.14472799830966526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,512,0.00871733327706655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,256,0.017456889152526855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,256,0.14042754968007407
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,3072,0.01627111103799608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,256,0.008298666940795051
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,2560,0.14560888873206243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,128,0.01686044368478987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,6144,64,0.006264888991912206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,6144,128,0.00833422193924586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,10240,0.20801866054534912
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,6144,32,0.006439111298984951
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,64,0.016530666086408827
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,6144,32,0.016541333662139047
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,768,0.14642932679918078
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,5120,65536,0.43114667468600804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,6144,256,0.007316444483068254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,5120,16384,0.12015199661254883
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,16384,0.05147288905249702
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,16384,0.22923734452989367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,6144,128,0.006631111105283101
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,5120,12288,0.09093599849277073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,16384,0.3143582344055176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,65536,0.8966266844007703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,12288,0.03992799917856852
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,6144,128,0.13712799549102783
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,12288,0.16339200072818333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,5120,8192,0.05832088655895657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,12288,0.24181156688266328
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,10240,0.1340364482667711
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,10240,0.03445777628156874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,5120,7168,0.055504000849194
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,8192,0.11086399687661065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,8192,0.17507733239067924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,7168,0.0984613365597195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,7168,0.02629511058330536
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,7168,0.15919555558098689
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,5120,6144,0.0454053348965115
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,65536,1.1526070700751412
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,65536,0.19043021731906465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,5120,5120,0.037463112009896174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,6144,0.023874666955735948
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,6144,0.14630844857957628
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,5120,4096,0.030834668212466772
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,5120,0.0736266639497545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,5120,0.021584888299306233
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,5120,0.14688977930280897
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,5120,3584,0.02738044493728214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,4096,0.05889066722657946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,4096,0.018414222531848483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,5120,10240,0.07600622044669257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,3584,0.0520702236228519
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,5120,3072,0.022261333134439256
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,3584,0.14529866642422146
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,3072,0.04700444473160637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,3072,0.015256888336605497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,3072,0.14776266945732966
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,10240,0.2082648939556546
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,8192,0.028838220569822524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,5120,2048,0.01695999999841054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,2560,0.014067555467287699
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,2560,0.041476445065604314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,2048,0.03535200158754984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,2048,0.14247555202907985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,5120,1536,0.013304000099500021
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,2048,0.012685333689053854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,6144,0.08653510941399468
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,1536,0.02959022257063124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,1536,0.010713777608341642
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,5120,1024,0.009558222360081142
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,1536,0.15044444137149388
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,4096,0.1401217778523763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,1024,0.022679110368092854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,5120,768,0.008943999807039896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,1024,0.14287466473049587
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,3584,0.016896888613700867
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,5120,512,0.007267555428875818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,768,0.009008000294367472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,768,0.1407617727915446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,5120,2560,0.019208000765906442
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,512,0.008657777474986183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,512,0.14314756128523085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,5120,256,0.006270222365856171
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,2560,0.14639643828074136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,256,0.016529776983790927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,5120,128,0.005544888890451855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,256,0.008099555969238281
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,256,0.14366843965318468
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,5120,64,0.0052382221652401825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,128,0.008037333687146505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,128,0.016134222348531086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,5120,128,0.14067110750410292
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,5120,32,0.005215999980767568
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,64,0.015839111473825242
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,32,0.015788444214397006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,4096,65536,0.3297528955671522
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,5120,1024,0.009350222018029954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,4096,16384,0.09428266684214275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,65536,0.14955555068122015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,768,0.02030311193731096
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,16384,0.040638221634758845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,4096,12288,0.07002400027381049
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,16384,0.19456888569725883
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,5120,512,0.017895110779338412
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,16384,0.3057182100084093
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,4096,10240,0.05754844347635905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,12288,0.03215288784768846
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,12288,0.1436026626163059
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,4096,8192,0.048878222703933716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,12288,0.2161511050330268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,10240,0.12167910734812419
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,10240,0.02759644389152527
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,4096,7168,0.041837334632873535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,10240,0.19157066610124376
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,8192,0.023226666781637404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,8192,0.09943288564682007
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,8192,0.15829066435496011
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,4096,6144,0.03625422385003831
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,7168,0.08905422025256687
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,7168,0.02195022172398037
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,4096,5120,0.029913776450686987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,65536,1.0217804378933377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,6144,0.019694222344292533
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,6144,0.14436977439456516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,6144,0.07799822092056274
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,4096,4096,0.02520711057715946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,5120,0.06669955783420138
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,5120,0.15099555916256374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,5120,0.01789866718981001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,4096,0.1413146654764811
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,4096,0.05388711227311028
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,4096,3072,0.019710222880045574
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,3584,0.048078222407235034
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,3584,0.013842666314707862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,3584,0.14276799890730116
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,4096,2560,0.01741422216097514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,3072,0.042846222718556724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,3072,0.012777778009573618
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,3072,0.15219910939534506
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,4096,2048,0.015365334020720588
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,65536,0.8325715594821506
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,2560,0.038374221987194486
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,2560,0.1453360054228041
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,2560,0.011522666447692446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,4096,1536,0.01240799989965227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,2048,0.010209778116808997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,2048,0.03207111027505662
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,2048,0.14730755488077799
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,4096,1024,0.010063110954231685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,1536,0.02678044471475813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,1536,0.1450559960471259
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,4096,768,0.008265777594513362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,1536,0.009350222018029954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,1024,0.021016000045670405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,1024,0.008416888614495596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,1024,0.15074666341145834
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,7168,0.14750399854448107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,4096,512,0.006667555620272954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,768,0.01897333396805657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,768,0.14308533403608534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,768,0.008001777860853408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,512,0.007641777396202087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,512,0.01683999929163191
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,512,0.1439911127090454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,4096,128,0.004934222333961063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,4096,3584,0.02244088881545597
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,4096,0.015152000718646579
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,256,0.007276444799370236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,4096,64,0.004562666846646203
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,128,0.01518311103185018
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,4096,128,0.007335999773608313
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,128,0.1390275557835897
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,4096,32,0.004906666775544484
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,64,0.015423110789722867
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,32,0.015141333142916361
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3584,65536,0.31563109821743435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3584,16384,0.08218044704861112
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,65536,0.14365243911743164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,16384,0.1963742176691691
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,4096,256,0.005581333405441708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,16384,0.27877510918511283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3584,12288,0.06405511167314318
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,16384,0.041153778632481895
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,4096,256,0.01583377851380242
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,4096,256,0.14621244536505804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3584,10240,0.058463109864128955
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,12288,0.14133421579996744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,12288,0.03166044420666165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,12288,0.21673066086239287
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3584,8192,0.0432951119210985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,10240,0.027411555250485737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,10240,0.1862106720606486
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,65536,1.0183511310153537
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3584,7168,0.04009688893953959
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,8192,0.023294222023752
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,8192,0.09762399726443821
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3584,6144,0.0343466666009691
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,7168,0.021336888273557026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,7168,0.14905511008368597
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,6144,0.07612622446484037
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,6144,0.14040356212192112
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3584,5120,0.029380444023344252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,6144,0.019693333241674636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3584,4096,0.02462755474779341
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,5120,0.01758666667673323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,5120,0.06444978051715426
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,5120,0.14065778255462646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3584,3584,0.02235200007756551
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,4096,0.052183998955620654
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,65536,0.7855644226074219
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,4096,0.015070221490330167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,4096,0.15105422337849936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3584,3072,0.019878221882714164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,3584,0.046837333175871104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,3584,0.14460710684458414
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3584,2560,0.01758044461409251
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,3072,0.012422222230169507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,2560,0.03751822312672933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,2560,0.14175200462341309
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,10240,0.11611999405754937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3584,2048,0.015254222684436373
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,2560,0.011390222443474663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,2048,0.03149155444569058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,2048,0.14149067136976454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,2048,0.010269333091047075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,7168,0.08760977453655666
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,1536,0.009296889106432596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,1536,0.026022222306993272
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,1536,0.1462399959564209
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3584,1024,0.009694221946928237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,1024,0.02089777754412757
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,1024,0.008674666285514832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3584,768,0.007968000239796108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,768,0.01860000027550591
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,768,0.1408204502529568
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3584,512,0.0069146665434042616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,3072,0.04218399855825636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,768,0.007984889050324758
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,3072,0.14641066392262778
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3584,256,0.00566133318675889
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,512,0.01647822227742937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,512,0.007651555869314406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,512,0.15140088399251303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3584,128,0.004951111144489712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,256,0.01517511076397366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,256,0.0075351107451650835
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,128,0.014793778459231058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3584,1536,0.0124177775449223
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3584,64,0.00490044429898262
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,128,0.007302222152551015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3584,32,0.005569777968857024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,64,0.0148053334818946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3584,32,0.01443377799457974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3072,65536,0.24325421121385363
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,1024,0.14578399393293592
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3072,16384,0.0722035566965739
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3584,3584,0.013757333159446716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,65536,0.13939732975429958
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,16384,0.1832204394870334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,16384,0.038872000243928694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3072,12288,0.05341955688264635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,65536,0.7355262438456217
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,16384,0.27936355272928876
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,8192,0.1570604509777493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,12288,0.13416888978746203
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3072,10240,0.046839998828040234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,12288,0.030524445904625788
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,12288,0.21512177255418566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,128,0.13957956102159289
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3072,8192,0.037618666887283325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,10240,0.02694844371742672
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,10240,0.11317600144280328
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,8192,0.09298133187823826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3072,7168,0.032440000110202365
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,8192,0.15590666400061712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,8192,0.022809776994917128
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3072,6144,0.030293332205878362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,7168,0.02100444502300686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,7168,0.14980622132619223
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,7168,0.08283999893400404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3072,5120,0.02374222212367588
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,6144,0.07285333342022367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,6144,0.019253333409627277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,6144,0.14596800009409586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3072,4096,0.020120888948440552
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,5120,0.061457779672410756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,5120,0.01720266706413693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,5120,0.14486311541663274
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3072,3584,0.018399111098713346
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,4096,0.0500257776843177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,4096,0.014530667000346713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,4096,0.1475493378109402
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3072,3072,0.016551999582184684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3584,256,0.14240622520446777
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,3584,0.044807109567854136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,3584,0.14458399348788792
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3072,2560,0.014674666855070325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,3072,0.03996444410747952
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,3072,0.012303111453851065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,2560,0.0355022218492296
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,2560,0.011023999916182624
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3072,2048,0.012753778033786349
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,65536,1.0105359819200304
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,2048,0.02996444371011522
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,2048,0.010002666877375709
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,2048,0.1443235609266493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3072,1536,0.009753777749008602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,1536,0.025358221597141687
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,1536,0.00903733323017756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,1536,0.14465067121717665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3072,1024,0.007624000310897827
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3072,768,0.006550222221348021
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,1024,0.00831200016869439
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,1024,0.019916445016860962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,1024,0.14450221591525608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3072,512,0.006051555689838197
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,768,0.01812799937195248
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,768,0.00794311116139094
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,768,0.1444800032509698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3072,256,0.004892444445027245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,512,0.14461333221859402
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,3584,0.013388444152143268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,512,0.0075937774446275495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3072,128,0.004531555705600315
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,256,0.014819555812411837
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,256,0.007313778003056844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,256,0.1433617803785536
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3072,64,0.0041786668201287585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,10240,0.18480889002482095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,3072,32,0.0041617775956789655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,3072,128,0.007309333317809635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,128,0.13685688707563612
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,64,0.014136888914638095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,32,0.01407999959256914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2560,65536,0.16735288831922743
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2560,16384,0.057562669118245445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,65536,0.13749333222707114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,16384,0.1734808948304918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,65536,0.6922737757364908
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,16384,0.037348445918824934
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,16384,0.2737209002176921
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2560,12288,0.04253866606288486
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,12288,0.030087110069062974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,3072,0.1396391126844618
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,12288,0.1273368861940172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2560,10240,0.03569066524505615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,12288,0.21186933252546522
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2560,8192,0.03054844339688619
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,3072,2560,0.1493368943532308
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,10240,0.02583377725548214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,10240,0.10550133387247722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,10240,0.1818915605545044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,65536,1.0058960384792752
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2560,7168,0.02555111050605774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,8192,0.08785511387719049
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,8192,0.15599555439419216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2560,6144,0.022613333331214056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,7168,0.0203813330994712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,7168,0.14438666237725153
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2560,5120,0.02009600069787767
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,6144,0.06861422459284465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,6144,0.14591288566589355
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,6144,0.01845066746075948
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,5120,0.05753155549367269
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,5120,0.016521778371598985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,5120,0.1488915549384223
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2560,4096,0.016146666473812528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,4096,0.013953778478834363
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,512,0.016174222032229107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,4096,0.046896000703175865
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2560,3584,0.015014222926563688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,3584,0.04241066508822971
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,3584,0.013025778035322825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,3584,0.14091289043426514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2560,3072,0.013445333474212222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,3072,128,0.014478221535682678
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2560,2560,0.011734222372372946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,3072,0.1405715545018514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,3072,0.012091555529170565
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,8192,0.021964444054497614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,2560,0.010888000329335531
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2560,2048,0.010281778044170803
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,7168,0.07878577709197998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,2048,0.027838223510318335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,2048,0.14605422814687094
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2560,1536,0.008373333348168267
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,2048,0.009720889230569204
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,1536,0.009001777403884465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,1536,0.02414400047726101
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,1536,0.14695999357435438
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2560,1024,0.006940444310506185
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,1024,0.019133332702848647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,4096,0.14179200596279568
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2560,768,0.005923555543025334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,1024,0.1467733383178711
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,768,0.017287999391555786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,768,0.007931555310885111
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,768,0.14579199420081243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2560,512,0.00518488883972168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,512,0.015524443652894763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,512,0.007628444168302748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2560,256,0.004523555437723796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,2560,0.03372977839575873
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,256,0.014452444182501899
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,2560,0.1420088873969184
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2560,128,0.004224888980388641
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,256,0.1466088824801975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,256,0.007027555671003129
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2560,64,0.0038542221817705366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,128,0.013792888985739814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,128,0.007236444287829929
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2560,32,0.0038844446341196695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,64,0.013737777868906656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,32,0.013746666411558786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2048,65536,0.17477777269151476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2048,16384,0.04557866520351834
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2560,1024,0.008303999900817871
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,65536,0.09320800171958075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,16384,0.16246399614546034
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,16384,0.02676888969209459
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2048,12288,0.03710044423739115
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,16384,0.24287109904819063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2560,3072,0.03801155421468947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,512,0.14589777257707384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,12288,0.02148888839615716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2048,10240,0.0321724447939131
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,12288,0.18753600120544434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2048,8192,0.025198222862349615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,10240,0.019101333287027147
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,10240,0.09841422239939372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,10240,0.16088088353474936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2048,7168,0.022879999544885423
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,65536,0.8693155712551541
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,8192,0.08182489209704928
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,8192,0.016492444607946608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,8192,0.14695466889275446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2048,6144,0.020424889193640817
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,7168,0.07319288783603244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,7168,0.015261333849694995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,7168,0.1426746712790595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2048,5120,0.01719377769364251
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,6144,0.0636977752049764
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,6144,0.14423288239373103
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,6144,0.014135999812020196
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2048,4096,0.014696889453464083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,5120,0.0537484453784095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,5120,0.1402453316582574
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,5120,0.0129466669427024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2048,3584,0.013388444152143268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,4096,0.043707556194729276
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,4096,0.011335999601417117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,4096,0.14387022124396429
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2048,3072,0.012065778175989786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,3584,0.03979377614127265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,3584,0.14009066422780356
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,3584,0.01074399964676963
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2048,2560,0.01071999967098236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,3072,0.035113778379228376
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,3072,0.009677333136399588
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,65536,0.6511715783013238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,3072,0.14566666550106472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2048,2048,0.009301332963837517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,12288,0.11934222115410699
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,2560,0.031199110878838435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,2560,0.14108178350660536
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2048,1536,0.007582222421964009
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,2048,0.026901332868470088
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,2048,0.008617777791288164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2048,1024,0.006230222268237009
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,1536,0.023017777336968318
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,1536,0.007920888562997183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,1536,0.150964445537991
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2048,768,0.005577777822812398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2560,128,0.13506578074561224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,1024,0.018244443668259513
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,1024,0.007335111498832703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2048,512,0.004878222114510006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,768,0.016532444291644625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,768,0.007270221908887227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,768,0.14290844069586858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2048,256,0.004243555582231945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,512,0.015138667490747241
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,512,0.006803555621041193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,512,0.1490106715096368
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2048,128,0.003859555555714501
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,256,0.01389955480893453
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,256,0.1435599989361233
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,128,0.013783999615245394
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,128,0.13524622387356228
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,128,0.006355555521117316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2048,32,0.0036231109665499795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,64,0.013395555317401886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,2560,0.009341333475377824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,2048,0.14393688572777644
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1536,65536,0.1491911146375868
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,2048,1024,0.14022666878170437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,65536,0.09039644400278728
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1536,16384,0.03898400068283081
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,65536,0.6089573436313206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,16384,0.15328621864318848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1536,12288,0.028792887926101685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,16384,0.2409022119310167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,2048,256,0.006582222051090664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,2048,64,0.00351555562681622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,12288,0.11169421672821045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,12288,0.020823111136754353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,12288,0.18605421649085152
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1536,10240,0.02514222264289856
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,65536,0.8588764402601453
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1536,8192,0.02035555574629042
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,10240,0.01905066602759891
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,2048,32,0.013131555583741931
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,10240,0.15942488776312932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,10240,0.09386310974756877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1536,7168,0.017907554904619854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,8192,0.1500693294737074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,8192,0.01611288885275523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1536,6144,0.016324443949593436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,7168,0.1432408889134725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,7168,0.014874666929244995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1536,5120,0.013839999834696451
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,6144,0.05729955434799194
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,6144,0.14925155374738905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,6144,0.013794666363133324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1536,4096,0.01223288890388277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,5120,0.05014933480156792
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,5120,0.012598222328556908
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,5120,0.1397040022744073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1536,3584,0.011181332998805575
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,4096,0.040998223755094744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,4096,0.011063111325105032
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1536,3072,0.010064000056849586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,3584,0.03756355577045017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,16384,0.025940444734361436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,3584,0.010426667001512317
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,3584,0.14298310544755724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1536,2560,0.008965333302815756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,3072,0.03295822276009454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,3072,0.009690666364298927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,3072,0.1405644416809082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1536,2048,0.00795288880666097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,2560,0.009015999734401703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,2560,0.1478782229953342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1536,1536,0.006913777854707506
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,2048,0.025179555018742878
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,2048,0.008427555362383524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,8192,0.07687822315427992
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,1536,0.021382222572962444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1536,1024,0.005555555638339784
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,1536,0.13973155286577013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,7168,0.06856089168124728
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1536,768,0.0052782222628593445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,1024,0.017290666699409485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,1024,0.0074311114019817775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,1024,0.1429093281428019
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1536,512,0.004519111166397731
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,768,0.016010666886965435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,768,0.007043555378913879
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,768,0.14729866716596815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,512,0.014488889111412896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,512,0.006639110959238476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,4096,0.14411821630265978
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,256,0.013571555415789286
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,256,0.1398764451344808
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1536,128,0.0038862224254343244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,256,0.00656355544924736
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,2560,0.029719110992219713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,128,0.013096888860066732
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,128,0.13617422845628527
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,128,0.006559111177921295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,2048,0.14503378338283962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1536,32,0.0038782221575578055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,64,0.012823999755912356
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1536,1536,0.007968000239796108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1536,32,0.01277511070171992
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1024,65536,0.10805510812335545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1024,16384,0.033231112692091204
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,65536,0.08891199694739448
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1536,256,0.004222222086456087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1536,512,0.14525866508483887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,16384,0.1424515512254503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1024,12288,0.024531554844644334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,65536,0.562754684024387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,16384,0.02533600065443251
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1024,10240,0.021352888809310064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,12288,0.0203840004073249
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,12288,0.18326132827334932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,10240,0.08695199754503037
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,65536,0.8455564181009928
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1024,8192,0.017542221479945712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,10240,0.15839378039042154
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1536,64,0.003839111162556542
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1024,7168,0.01610311037964291
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,8192,0.015506666567590503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,8192,0.07110044691297743
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,8192,0.14760177665286595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1024,6144,0.014083556003040738
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,7168,0.06020977762010363
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,7168,0.014311111635631986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,7168,0.14432266023423937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1024,5120,0.012460444536474017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,6144,0.054013331731160484
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,6144,0.1406408945719401
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,6144,0.013210666676362356
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1024,4096,0.01070755554570092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,5120,0.04621333214971754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,5120,0.012425777812798819
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,5120,0.14794845051235622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1024,3584,0.009711999860074785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,4096,0.03714933329158359
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,16384,0.23590400483873156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,4096,0.010783111055692038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,4096,0.14188710848490396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1024,3072,0.009005332986513773
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,3584,0.034867554903030396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,3584,0.01035733355416192
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,3584,0.1497982210583157
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,12288,0.10350844595167373
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1024,2560,0.007978666987684038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,3072,0.13967822657691106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,3072,0.031144890520307753
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,3072,0.009441778063774109
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1024,2048,0.006940444310506185
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1024,1536,0.006225777582989798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,2560,0.027427555786238775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,2560,0.008996444443861643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,10240,0.018419555491871305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1024,1024,0.005201777650250329
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,2048,0.023689778314696416
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,2048,0.14479555024041071
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,1536,0.020236444142129686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,1536,0.007984889050324758
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1024,768,0.004553777890072929
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,1536,0.15050221814049616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,1024,0.007286222444640265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,1024,0.016556445095274184
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,1024,0.14285422696007624
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,768,0.13969511455959743
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1024,512,0.0041911109454101985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,512,0.1423128843307495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,768,0.006952889263629913
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,512,0.014159111513031853
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,512,0.006559999866618051
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1024,128,0.0031973332580592898
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,256,0.00629333323902554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,256,0.14834845066070557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,128,0.012776888906955719
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1024,64,0.003156444471743372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,128,0.006597333484225803
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1024,32,0.0031582222630580268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,64,0.012706666356987424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,32,0.012138666378127204
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,768,65536,0.08249244425031874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,2560,0.14137510458628336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,1024,2048,0.008339555727110969
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,768,16384,0.027447111076778833
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,65536,0.08886222044626872
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,768,16384,0.14129777749379477
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,1024,256,0.003527111063400904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,768,0.015483554866578845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,16384,0.23511823018391928
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,768,12288,0.01960355540116628
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,16384,0.02496444516711765
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,1024,256,0.013112000293201871
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,768,10240,0.01684177749686771
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,12288,0.020000888241661918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,65536,0.8438337643941244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,1024,128,0.13651822672949895
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,768,8192,0.014454222387737699
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,10240,0.017903111047214933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,10240,0.1570382250679864
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,8192,0.01512977812025282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,8192,0.14200977484385172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,768,7168,0.01316088851955202
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,768,7168,0.06341689162784152
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,7168,0.14763732751210532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,768,6144,0.011686222420798408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,7168,0.014296889305114746
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,768,6144,0.05403555432955424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,768,65536,0.5576151212056478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,6144,0.1402133305867513
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,768,5120,0.010639111200968424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,6144,0.01311644415060679
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,768,12288,0.10289066367679173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,5120,0.15233777629004583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,5120,0.012115555504957834
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,768,4096,0.009001777403884465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,768,3584,0.008300444318188561
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,768,10240,0.08619378010431926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,768,4096,0.03848888807826572
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,4096,0.14819910791185167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,768,3072,0.007485333416197035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,768,3584,0.034576889541414045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,3584,0.010052444206343757
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,768,8192,0.0705946683883667
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,3584,0.14242045084635416
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,768,2560,0.006619555668698416
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,768,3072,0.030580444468392268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,3072,0.009381333159075843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,3072,0.1430471075905694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,768,2048,0.006302222195598815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,768,2560,0.02773333258099026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,2560,0.008962666822804345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,2560,0.14509067270490858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,768,1536,0.005573333137565189
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,768,2048,0.023707555400000677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,2048,0.008282666405042013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,768,1024,0.004528000122971005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,1536,0.14677600065867105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,768,1024,0.016492444607946608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,1024,0.1467155615488688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,768,768,0.0041848888827694785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,12288,0.18507111072540283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,1024,0.007256888680987888
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,768,0.0069208890199661255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,768,768,0.015098666151364645
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,768,0.14096532927619085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,4096,0.010753778119881948
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,768,512,0.013774221969975365
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,512,0.006649777707126405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,512,0.14212177859412298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,768,256,0.003559111307064692
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,768,256,0.013043555948469373
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,256,0.006299555715587404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,256,0.14099821779463026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,768,128,0.01238044434123569
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,128,0.13209244940016004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,768,1536,0.020231111182106864
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,1536,0.007667555577225155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,768,128,0.006281777802440856
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,768,32,0.0033368888414568375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,768,64,0.01236533373594284
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,512,65536,0.05756533145904541
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,768,32,0.012106666962305704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,768,5120,0.04515022370550367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,512,16384,0.018966222802797954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,65536,0.08885866403579712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,768,512,0.0038933331767717996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,512,16384,0.1313448879453871
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,16384,0.02437066700723436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,16384,0.2346471150716146
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,768,128,0.0031964443624019623
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,512,12288,0.0943982203801473
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,12288,0.18023199505276152
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,512,10240,0.01612977849112617
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,65536,0.8381910853915744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,12288,0.019588443968031142
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,768,2048,0.15153333875868055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,512,8192,0.013781333135233985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,768,64,0.0034862222770849862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,512,10240,0.07983199755350749
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,10240,0.15775377220577663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,512,7168,0.01238666640387641
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,512,8192,0.06373511420355903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,8192,0.14465510845184326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,512,6144,0.011064888702498542
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,512,7168,0.054980443583594434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,7168,0.1468284395005968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,7168,0.014101333088344999
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,512,5120,0.009703999592198266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,6144,0.01312444441848331
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,512,6144,0.049976888630125255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,6144,0.14188978407118055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,512,5120,0.042488889561759106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,5120,0.012142221960756512
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,512,4096,0.008653333617581261
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,512,65536,0.5166426764594184
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,512,4096,0.03436266713672214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,4096,0.1499884393480089
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,512,3584,0.00793422261873881
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,4096,0.010679110884666443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,512,3584,0.03193511234389411
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,3584,0.14250132772657606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,512,3072,0.007299555672539606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,10240,0.017609778377744887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,512,3072,0.0281333327293396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,3072,0.1530239979426066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,512,2560,0.006604444649484422
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,8192,0.015234667393896313
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,3072,0.009394666386975182
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,512,2048,0.005541333307822545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,512,2560,0.025607999828126695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,2560,0.008957333034939235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,512,2048,0.022299556268586054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,2048,0.008314666648705801
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,2048,0.14978933334350586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,512,1536,0.004877333425813251
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,5120,0.14672533671061197
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,1536,0.007669332954618666
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,512,1024,0.0041982221106688184
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,1536,0.15077688958909777
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,512,12288,0.018252443936136033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,512,768,0.0041964443193541635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,512,1024,0.015850666496488783
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,1024,0.007281777759393056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,1024,0.14756800068749323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,512,512,0.0037084445357322693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,512,768,0.014757333530320061
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,768,0.0069351109365622205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,768,0.14505955908033583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,512,256,0.0035297779573334586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,512,512,0.013404444687896304
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,512,0.006578666468461354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,512,0.14973510636223689
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,512,128,0.0029102222373088202
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,512,256,0.012638222012254926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,256,0.0062773335311147906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,256,0.14422755771213108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,512,64,0.002833777831660377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,512,128,0.012082666986518435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,512,32,0.002809777855873108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,128,0.00628355559375551
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,256,65536,0.037794666157828435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,512,64,0.012072000238630505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,512,32,0.011740444435013665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,2560,0.14059200551774767
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,256,16384,0.01422044469250573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,65536,0.08832533491982354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,256,16384,0.13066221608055964
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,16384,0.024013333850436743
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,16384,0.2309039963616265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,256,12288,0.01198755535814497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,512,1536,0.019208888212839764
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,256,65536,0.5158915519714355
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,256,10240,0.012763555679056378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,12288,0.019381332728597853
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,256,12288,0.09434844387902154
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,12288,0.18458222018347847
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,256,8192,0.013014222184816996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,256,10240,0.07905511061350505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,10240,0.017510222064124215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,10240,0.15776444805992976
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,65536,0.8366142378913032
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,512,3584,0.009963555468453301
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,256,8192,0.061443554030524365
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,8192,0.01513777838812934
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,256,6144,0.010695111420419482
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,256,7168,0.056417776478661426
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,7168,0.1463128858142429
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,7168,0.014139556222491793
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,256,5120,0.009329777624871995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,256,6144,0.048024889495637685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,6144,0.15042577849494085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,6144,0.01309866706530253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,256,4096,0.00829777783817715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,256,5120,0.04071199893951416
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,5120,0.1478239960140652
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,5120,0.012183999849690331
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,256,3584,0.007253333098358578
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,256,4096,0.033589333295822144
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,4096,0.010661332971519895
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,4096,0.1495928896798028
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,256,3072,0.006607111129495833
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,256,3584,0.030894223186704848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,3584,0.010002666877375709
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,3584,0.145123561223348
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,256,2560,0.006250666661394968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,256,3072,0.027775110469924078
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,3072,0.14479288789961073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,3072,0.009331555830107795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,256,2048,0.00555288874440723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,256,2560,0.025245333711306255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,2560,0.008769777913888296
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,2560,0.14415021737416586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,256,1536,0.004578666554556953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,256,2048,0.022016889519161646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,2048,0.14538933171166316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,256,7168,0.011734222372372946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,256,1536,0.01923466722170512
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,256,1024,0.004243555582231945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,1536,0.1454533338546753
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,8192,0.14529066615634495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,256,768,0.0038728887836138406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,256,1024,0.01572444372706943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,1024,0.007272000114123027
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,256,768,0.014440889159838358
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,256,512,0.0034959999223550162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,768,0.0069466663731469055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,768,0.14540089501274958
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,512,0.006624889042642381
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,256,512,0.013068444199032254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,512,0.14492621686723497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,256,256,0.0031848889258172777
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,256,256,0.012100444071822695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,256,0.006296000132958095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,256,128,0.00286666676402092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,256,0.14419289429982504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,128,0.006590222318967183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,256,128,0.012102222277058495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,128,0.13786755667792425
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,256,32,0.0028275555620590844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,2048,0.00813422186507119
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,256,64,0.01204177737236023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,128,65536,0.030828442838456895
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,256,1536,0.007634667058785756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,256,1024,0.14157244894239637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,65536,0.08818044265111287
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,128,16384,0.01071733319097095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,512,128,0.1335973342259725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,128,65536,0.5157484478420681
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,16384,0.023927110764715407
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,128,16384,0.1299662192662557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,128,12288,0.012759110993809171
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,12288,0.01953066719902886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,128,10240,0.011396444506115384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,12288,0.18315111266242134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,256,64,0.0028400001012616684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,10240,0.017529777354664274
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,10240,0.15573510858747694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,128,8192,0.0101742222905159
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,256,32,0.011715555356608497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,8192,0.014931556251313953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,8192,0.14636622534857854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,128,8192,0.06347911225424872
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,128,7168,0.009665778113736046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,128,6144,0.009139555195967356
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,7168,0.014110222458839417
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,128,7168,0.0554328891966078
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,7168,0.14751022391849095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,128,5120,0.008285333712895712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,128,6144,0.048226667775048145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,6144,0.013079110946920184
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,6144,0.14368532763587102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,128,4096,0.007497777541478475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,128,5120,0.04056533177693685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,5120,0.1442293326059977
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,5120,0.012061333490742577
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,16384,0.228965335422092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,128,3584,0.006958222223652734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,4096,0.010692444112565784
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,128,4096,0.0333591103553772
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,4096,0.1458079947365655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,128,12288,0.09427911043167114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,128,3072,0.006266666783226862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,128,3584,0.03090488910675049
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,3584,0.1455084482828776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,65536,0.8367617925008138
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,128,2560,0.005767999837795894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,3584,0.010010666317409938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,3072,0.009311999711725447
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,128,3072,0.027621333797772724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,3072,0.1434968842400445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,128,2560,0.025387555360794067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,128,10240,0.07843999730216132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,2560,0.15193955103556314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,128,1536,0.004556444370084339
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,128,2048,0.02196088929971059
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,2048,0.1465946700837877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,128,1024,0.0038568890757030914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,128,1536,0.019183999962276883
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,1536,0.1526026725769043
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,128,1024,0.015812445018026564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,1024,0.007247111035717859
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,1024,0.1447475618786282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,128,768,0.014450665977266101
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,768,0.006903111106819577
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,128,512,0.003272888975010978
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,128,512,0.01311822235584259
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,512,0.1483057737350464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,128,256,0.003156444471743372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,512,0.006578666468461354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,128,256,0.012084444363911947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,256,0.1442293326059977
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,128,2048,0.0052639999323421055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,128,128,0.002862222285734283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,2560,0.008678221868144141
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,256,0.006265777680608962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,2048,0.00832711077398724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,128,128,0.011705777711338468
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,128,64,0.002505777817633417
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,128,0.00628977765639623
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1024,128,1536,0.007635555333561367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,128,32,0.0028453332682450614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,64,65536,0.03176000052028232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,128,64,0.011702222128709158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,128,768,0.0038728887836138406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,64,16384,0.010028444230556488
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,128,32,0.011763555308183035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,64,12288,0.00866844422287411
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,64,10240,0.009020444419648912
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,64,8192,0.00831733312871721
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,64,16384,0.12954044342041016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,768,0.14390667279561362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,64,10240,0.07778844568464491
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,64,7168,0.007948444121413762
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,64,6144,0.007292444507280986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,64,5120,0.008027555214034187
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,64,65536,0.5152159796820747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,64,7168,0.05550666650136312
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,64,6144,0.047319110896852284
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,64,4096,0.007328000333574083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,64,3584,0.006974221931563483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,64,5120,0.03983822133806016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,64,3072,0.006231110956933763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,64,2560,0.005708444449636672
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,64,4096,0.03340977761480544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,64,2048,0.005208000126812193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,64,3072,0.027689778142505225
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,64,2560,0.025062221619817946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,64,1536,0.004559999952713649
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,64,1024,0.0038720000949170855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,64,1536,0.018936889039145578
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,64,768,0.003559111307064692
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,64,1024,0.015495111544926962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,64,512,0.0034684443639384378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,64,256,0.0031715554909573663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,64,768,0.014205333259370593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,64,128,0.0028764444092909494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,64,512,0.012804444465372296
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,64,64,0.002874666617976295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,64,256,0.012048000262843238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,64,32,0.0028106667515304354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,64,128,0.011688888900809817
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,64,12288,0.09356977542241414
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,64,64,0.011405333048767514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,64,8192,0.06364711125691731
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,32,16384,0.012401777837011548
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,64,32,0.011412444214026133
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,32,12288,0.01036711119943195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,32,10240,0.009687110781669617
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,32,16384,0.1295822196536594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,32,12288,0.09379288885328506
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,32,8192,0.009346666435400644
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,32,10240,0.07818844583299425
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,32,7168,0.0086666668454806
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,64,3584,0.030825777186287776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,32,6144,0.00831733312871721
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,32,5120,0.007622222105662028
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,32,8192,0.06363377968470256
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,32,4096,0.007642666498819987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,32,7168,0.05469333463244968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,32,6144,0.047446221113204956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,32,3584,0.007273777491516537
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,32,3072,0.006504888749784893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,32,5120,0.03925333420435587
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,32,2560,0.0058746664888328975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,1024,128,128,0.13931555218166775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,32,3584,0.03087377879354689
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,32,2048,0.005598222215970357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,32,3072,0.027726221415731642
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,32,1536,0.004895110925038655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,32,2560,0.025069332785076563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,32,1024,0.004214222232500712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,32,2048,0.02164977788925171
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,32,1536,0.018924444913864136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,32,1024,0.015493333339691162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,32,768,0.014441778262456259
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,32,256,0.0031439999325407874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,32,512,0.013105777402718862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,32,128,0.002809777855873108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,32,256,0.012088889049159156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,32,64,0.0028231110837724474
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,32,128,0.01202933324707879
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,32,32,0.002805333377586471
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,32,64,0.011722666521867117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,32,65536,0.5148497687445747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,32,32,0.011696889168686338
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,64,2048,0.02197866638501485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1024,32,4096,0.03311377763748169
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,32,768,0.0038782221575578055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,32,512,0.0035004446076022256
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1024,32,65536,0.026945778065257605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,65536,16384,0.8801795641581217
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,16384,0.33965155813429093
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,65536,12288,0.6488782034979926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,65536,16384,1.095465766059028
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,65536,12288,0.792960908677843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,65536,10240,0.616871092054579
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,12288,0.24697865380181205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,65536,10240,0.682520866394043
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,10240,0.2187217738893297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,65536,8192,0.4638551076253255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,65536,8192,0.5311751365661621
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,8192,0.1783777740266588
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,65536,7168,0.4573991033766005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,12288,0.7876044379340278
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,16384,1.0256648593478732
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,10240,0.61899291144477
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,65536,7168,0.455093330807156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,65536,6144,0.3918906582726373
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,7168,0.15529333220587835
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,8192,0.5102764235602485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,65536,6144,0.42647555139329696
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,6144,0.13639110989040798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,65536,5120,0.3806408776177301
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,65536,4096,0.22465599907769096
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,7168,0.45243554645114475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,5120,0.11755555205874973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,65536,4096,0.2667457792494032
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,6144,0.3952942159440782
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,65536,3584,0.23493422402275932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,4096,0.09729244311650594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,5120,0.3386711014641656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,65536,3584,0.2431404325697157
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,65536,3072,0.1962897777557373
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,3584,0.08586311340332031
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,65536,3072,0.20498932732476127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,4096,0.3027119901445177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,3072,0.07716088824801974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,65536,2560,0.1616062190797594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,3584,0.25384177101982963
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,65536,2560,0.17782755692799887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,65536,2048,0.12301954958173965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,3072,0.23119465510050455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,2560,0.06711288955476549
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,2048,0.05707199705971611
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,65536,1536,0.09028889073265924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,65536,2048,0.16280089484320745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,2048,0.17193688286675346
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,2560,0.19782311386532256
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,65536,1536,0.11670400036705865
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,65536,5120,0.33589333958095974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,65536,1024,0.06571377648247613
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,65536,768,0.05441155698564318
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,65536,1024,0.09373778104782104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,1536,0.14468355973561606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,1024,0.14407732751634386
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,65536,512,0.03908088803291321
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,65536,768,0.07339911328421699
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,768,0.14799022674560547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,768,0.0332275562816196
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,65536,256,0.03165155649185181
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,65536,512,0.06451733244789971
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,512,0.1421040031645033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,512,0.030343112018373277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,65536,128,0.024757333927684363
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,65536,256,0.05912711222966512
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,256,0.028051555156707764
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,65536,64,0.021218667427698772
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,65536,128,0.055424888928731285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,65536,32,0.021532444490326777
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,65536,64,0.05477955606248644
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,65536,32,0.05468177795410156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,1536,0.04519911275969612
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,1024,0.03654222355948554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,16384,65536,0.9121066199408637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,16384,16384,0.2257031069861518
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,256,0.14399645063612196
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,65536,128,0.1437564425998264
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,65536,128,0.027730666928821143
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,16384,0.09690577454037136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,16384,65536,1.3383084403143988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,16384,16384,0.36622577243381077
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,16384,12288,0.21194310983022055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,65536,0.37436800532870823
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,16384,12288,0.24267376793755424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,16384,10240,0.15649599499172634
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,16384,0.3776666588253445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,16384,10240,0.20296888881259492
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,16384,8192,0.12808444764879015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,12288,0.2958915498521593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,10240,0.2665128972795275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,8192,0.05156800150871277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,16384,8192,0.16365333398183188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,16384,7168,0.1223564412858751
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,8192,0.21399554941389295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,16384,6144,0.09614577558305527
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,16384,7168,0.15448355674743652
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,7168,0.046153777175479464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,7168,0.1940444442960951
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,16384,5120,0.0858408874935574
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,16384,6144,0.12789866659376356
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,6144,0.04084177811940511
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,6144,0.17029155625237358
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,5120,0.03550488750139872
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,16384,4096,0.06944800085491605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,5120,0.1528986692428589
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,12288,0.07437333133485582
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,16384,4096,0.09046044614579941
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,16384,3584,0.0714320010609097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,4096,0.1589697731865777
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,10240,0.06206488609313965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,16384,3072,0.05115911033418444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,16384,3584,0.08123555448320177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,3584,0.14609866672092015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,3584,0.026212443908055622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,16384,2560,0.04822755522198147
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,16384,3072,0.07050489054785834
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,3072,0.02426044477356805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,16384,2560,0.06232355700598823
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,2560,0.021697777840826247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,16384,2048,0.039860443936453924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,2560,0.14216355482737222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,16384,2048,0.0528400010532803
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,2048,0.018785778019163344
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,2048,0.14388533433278403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,16384,1536,0.030327998929553564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,16384,5120,0.10937777492735122
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,16384,1024,0.021013332737816706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,16384,1536,0.04189777705404493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,1536,0.14339555634392634
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,4096,0.029314666986465454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,1024,0.013925333817799887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,16384,1024,0.03295822276009454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,1024,0.1434746715757582
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,16384,768,0.028584000137117174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,768,0.1462213330798679
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,768,0.012662221988042196
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,3072,0.14575200610690647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,512,0.14086577627393934
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,16384,256,0.01218933363755544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,512,0.011397333608733283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,16384,256,0.02295911146534814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,256,0.1478559970855713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,16384,128,0.010344888601038191
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,256,0.010771555205186209
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,16384,128,0.021730666359265644
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,128,0.011009777585665384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,128,0.14162577523125544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,16384,64,0.009694221946928237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,16384,32,0.010016000105275048
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,16384,768,0.01741066740618812
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,16384,32,0.021291555629836187
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,16384,65536,1.4660755793253581
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,16384,512,0.013574221895800697
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,16384,512,0.024569777978791132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,12288,16384,0.17585599422454834
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,12288,65536,0.6347671084933811
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,65536,0.2710079881880018
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,12288,16384,0.26362045605977374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,12288,12288,0.13814043998718262
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,16384,1536,0.016381333271662395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,16384,0.3080204327901204
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,12288,65536,1.0942506790161133
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,12288,12288,0.19777333736419678
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,12288,10240,0.12099556128184001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,16384,64,0.02141688929663764
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,12288,0.05540977915128072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,12288,8192,0.09003733264075385
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,12288,0.27182666460673016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,12288,10240,0.16506755352020264
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,10240,0.047372443808449634
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,10240,0.2080684502919515
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,12288,8192,0.13542221652136907
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,8192,0.038252443075180054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,12288,7168,0.12104800012376572
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,12288,6144,0.07682933410008748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,7168,0.03478666808870103
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,7168,0.16249333487616643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,12288,6144,0.10599555571873982
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,6144,0.030798223283555772
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,12288,5120,0.06575555271572538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,16384,0.0714062253634135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,5120,0.026664000418451097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,5120,0.1449706686867608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,65536,1.1790648566351998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,12288,4096,0.0758604473537869
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,4096,0.1457137796613905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,4096,0.023772444989946153
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,12288,3584,0.03882311118973626
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,12288,3584,0.06830844614240858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,3584,0.14586488405863443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,12288,7168,0.07405689027574328
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,12288,3072,0.04114044374889798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,8192,0.1785066657596164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,3072,0.018561777141359117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,12288,3072,0.05879555808173286
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,3072,0.14619199434916177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,12288,2560,0.03369333346684774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,12288,2048,0.027161777019500732
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,12288,2560,0.05187199844254387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,2560,0.14466222127278647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,6144,0.14437244998084173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,2560,0.016982222596804302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,12288,2048,0.04399466514587402
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,12288,1536,0.021271111236678228
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,2048,0.14409244060516357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,2048,0.015354666444990369
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,12288,5120,0.09134044249852498
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,12288,1024,0.015306666493415833
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,12288,1536,0.03513688842455546
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,1536,0.1469715568754408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,1536,0.013391111459996967
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,12288,768,0.012709333664841123
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,12288,1024,0.028543998797734577
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,1024,0.011384888655609555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,1024,0.14115999804602727
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,12288,512,0.010034666293197209
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,12288,768,0.024619556135601465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,768,0.14721777704026964
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,768,0.010327111515733931
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,12288,256,0.007640000018808577
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,12288,512,0.021401777863502502
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,512,0.14148444599575466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,512,0.009690666364298927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,3584,0.02011022302839491
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,12288,128,0.0069146665434042616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,12288,256,0.019903999235894945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,256,0.008967999782827165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,12288,64,0.0063582224150498705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,12288,32,0.0064693333374129395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,12288,128,0.019271110494931538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,12288,128,0.009353777600659264
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,128,0.139174222946167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,12288,64,0.018922666708628338
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,12288,32,0.01886755559179518
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,10240,16384,0.1521706713570489
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,12288,4096,0.04554399847984314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,10240,65536,0.5506942007276747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,65536,0.2507253222995334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,10240,16384,0.23518400722079805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,10240,12288,0.11993243959214951
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,16384,0.06690311431884766
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,10240,12288,0.1764977773030599
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,10240,65536,0.9894107182820638
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,16384,0.30395110448201496
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,12288,0.05229777759975857
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,12288,256,0.1458479960759481
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,12288,0.24713778495788574
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,10240,8192,0.07595555649863349
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,10240,10240,0.14897421995798746
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,10240,0.20686311191982695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,10240,8192,0.1229742235607571
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,8192,0.035642666949166193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,10240,7168,0.07731911208894518
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,10240,7168,0.10837066835827297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,7168,0.03239555491341485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,7168,0.16064443853166369
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,10240,6144,0.06819111108779907
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,10240,5120,0.052601777844958834
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,10240,6144,0.09496088822682698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,6144,0.02888710962401496
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,6144,0.15134933259752062
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,65536,1.1450506846110027
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,10240,4096,0.04417333337995741
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,10240,5120,0.0806915561358134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,5120,0.025403555896547105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,5120,0.1430062188042535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,10240,3584,0.039761778381135725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,4096,0.02167111138502757
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,10240,4096,0.06912799676259358
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,4096,0.14906133545769587
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,10240,3072,0.037137776613235474
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,10240,3584,0.06090844339794583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,3584,0.1423164473639594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,10240,0.04409511221779717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,10240,2560,0.029598222838507757
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,3072,0.017986666825082567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,3072,0.14790577358669704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,10240,2048,0.026174222429593403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,10240,2560,0.0454328887992435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,2560,0.14201066229078504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,2560,0.01648355523745219
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,10240,1536,0.02037244372897678
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,10240,2048,0.0398871103922526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,2048,0.14459199375576443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,2048,0.0148062225845125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,10240,1024,0.015374221735530429
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,10240,1536,0.03189511100451151
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,1536,0.1416106621424357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,10240,1024,0.026029333472251892
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,1024,0.010845333337783813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,1024,0.1450835598839654
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,10240,768,0.012797333300113678
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,10240,768,0.022709333234363135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,768,0.010213333699438307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,768,0.14647555351257324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,10240,512,0.009642666412724389
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,10240,10240,0.0926168892118666
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,10240,256,0.007621333003044128
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,10240,512,0.0199306673473782
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,512,0.00941955546538035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,3584,0.019681778219011094
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,10240,3072,0.05279022124078539
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,10240,256,0.018600889378123812
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,10240,128,0.0069546666410234236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,256,0.009007111191749573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,256,0.14145955774519178
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,10240,64,0.006562666760550604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,10240,128,0.01791466606987847
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,8192,0.17723555035061309
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,128,0.009025777379671732
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,128,0.14302399423387316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,10240,32,0.006583999842405319
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,10240,64,0.01755644381046295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,10240,32,0.01756088932355245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,10240,1536,0.012728888955381183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,8192,16384,0.11661867300669353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,8192,65536,0.43927910592820907
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,65536,0.2155768871307373
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,8192,16384,0.20435910754733613
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,8192,12288,0.09154844284057617
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,16384,0.05554844604598152
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,16384,0.2691635555691189
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,10240,512,0.14607821570502386
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,8192,10240,0.08066399892171223
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,12288,0.042747557163238525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,12288,0.21789956092834473
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,8192,10240,0.13091111183166504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,10240,0.18440888987647164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,8192,8192,0.10879821909798516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,8192,0.15889599588182238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,8192,7168,0.05783110857009888
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,8192,0.029674665795432195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,65536,1.0047476026746962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,8192,7168,0.09662133455276489
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,7168,0.14726666609446207
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,8192,6144,0.05508088734414843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,7168,0.027405333187845018
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,8192,12288,0.15495644675360784
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,8192,65536,0.8558168941073947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,6144,0.02477866742346022
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,6144,0.14503467082977295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,8192,5120,0.04654755526118808
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,8192,4096,0.03733333283000522
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,5120,0.021319111188252766
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,5120,0.14577599366505942
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,10240,0.03635111120012071
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,8192,8192,0.06794844733344184
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,8192,4096,0.059226665231916636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,8192,3584,0.033193777004877724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,4096,0.143248889181349
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,8192,3072,0.028296000427669946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,4096,0.01844800015290578
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,3584,0.017216889394654166
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,8192,3584,0.05350311266051399
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,3584,0.14222933186425105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,8192,2560,0.02515999972820282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,3072,0.015444444285498725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,8192,3072,0.04736977815628052
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,3072,0.14471199777391222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,8192,2048,0.020777778493033517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,8192,2560,0.04169333312246534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,2560,0.14154133531782362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,2560,0.014112000664075216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,8192,1536,0.01570044457912445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,8192,2048,0.035818666219711304
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,2048,0.012895111408498553
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,2048,0.14430044756995306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,8192,1024,0.01292711082432005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,8192,1536,0.029281778468026057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,1536,0.010859555668301053
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,8192,768,0.010808889236715106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,8192,1024,0.023686221904224817
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,1024,0.14393777317470974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,1024,0.009637333452701569
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,8192,512,0.009486222432719337
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,8192,768,0.020615999897321064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,768,0.009018667042255402
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,8192,512,0.018536888890796237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,8192,6144,0.08535555336210464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,512,0.1450444459915161
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,512,0.00848444468445248
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,8192,128,0.006650666809744305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,8192,256,0.01716088917520311
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,256,0.008079999850855933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,256,0.14673156208462185
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,8192,5120,0.07245333327187432
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,8192,64,0.006186666587988536
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,8192,32,0.006287999865081575
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,8192,128,0.008009778128729926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,128,0.13527199957105848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,8192,64,0.016492444607946608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,8192,32,0.0162444445821974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,7168,65536,0.4339786635504828
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,1536,0.14378400643666586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,65536,0.2015155553817749
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,7168,16384,0.12371822198232015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,8192,768,0.142922666337755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,7168,16384,0.1919377777311537
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,8192,256,0.007277333074145847
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,16384,0.0544435547457801
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,7168,12288,0.08873599767684937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,16384,0.27454400062561035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,7168,12288,0.1463680002424452
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,8192,128,0.016846223009957206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,12288,0.03960977660285102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,7168,10240,0.0944897797372606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,7168,8192,0.07562222083409627
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,10240,0.034097777472601995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,7168,10240,0.12467111481560601
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,10240,0.18346667289733887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,7168,7168,0.06954933537377252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,8192,0.02863911125395033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,8192,0.1577271090613471
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,65536,0.9943066702948676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,7168,0.02555199960867564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,7168,7168,0.09169689151975845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,7168,0.14578399393293592
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,7168,6144,0.06080089012781779
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,7168,5120,0.0477804442246755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,6144,0.023461333579487268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,7168,6144,0.08021866612964206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,6144,0.14657066928015813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,7168,65536,0.7954346868726941
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,7168,4096,0.037588444021013044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,7168,5120,0.06761599911583795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,5120,0.02089511189195845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,5120,0.14409511619144016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,7168,3584,0.033987555238935686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,7168,4096,0.05579821930991279
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,4096,0.14692889319525823
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,4096,0.01811288959450192
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,7168,3072,0.02867022156715393
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,12288,0.21360622511969674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,3584,0.016509332590632968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,3584,0.14419021871354845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,7168,2560,0.024667556087176006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,7168,3072,0.04460888769891527
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,3072,0.14534933037228054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,7168,2048,0.020807999703619216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,2560,0.014055111342006259
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,2560,0.14176532957288954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,7168,8192,0.1024071110619439
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,7168,1536,0.016872889465755887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,7168,2048,0.03424711028734843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,2048,0.012485333614879184
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,2048,0.14716088771820068
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,7168,1024,0.01147466649611791
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,7168,1536,0.027981334262424048
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,1536,0.14414310455322266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,1536,0.010605333579911126
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,7168,768,0.009436444275909001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,1024,0.009327111144860586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,1024,0.15016178290049234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,7168,512,0.007275555696752336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,7168,768,0.019890666007995605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,768,0.14401866330040827
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,768,0.008823110825485652
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,7168,256,0.006259555617968242
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,7168,512,0.017844445175594754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,512,0.14257866806454128
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,512,0.008360000120268928
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,7168,3584,0.050929778152041964
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,7168,128,0.005392888767851724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,7168,256,0.016521778371598985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,256,0.007987555530336168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,256,0.15049689345889622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,3072,0.015227556228637695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,7168,64,0.004912000149488449
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,7168,128,0.00795999997191959
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,7168,128,0.1360337734222412
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,7168,32,0.00490044429898262
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,7168,2560,0.03935111231274075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,7168,64,0.015860444969601102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,7168,32,0.015831111205948722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,6144,16384,0.09509244230058457
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,6144,65536,0.338626676135593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,65536,0.15158755249447292
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,6144,16384,0.17435910966661242
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,7168,1024,0.022276444567574397
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,6144,12288,0.07848888635635376
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,16384,0.23327999644809297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,16384,0.04057333204481337
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,6144,10240,0.05742488967047798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,6144,12288,0.13260355260637072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,12288,0.03143999973932902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,6144,65536,0.7381342252095541
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,12288,0.1877511077457004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,6144,8192,0.0544488893614875
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,6144,10240,0.11491910616556804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,10240,0.0269813338915507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,6144,7168,0.04574755496448941
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,8192,0.02292977770169576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,6144,8192,0.0934915542602539
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,8192,0.14992445045047337
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,6144,7168,0.08340266678068374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,6144,6144,0.03361422154638503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,7168,128,0.016376889414257474
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,7168,0.1458915604485406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,7168,0.02088711162408193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,6144,6144,0.07323022021187676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,6144,0.14303110705481634
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,6144,5120,0.06118933359781901
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,5120,0.14182755682203504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,6144,4096,0.027406222290462915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,5120,0.01715555621518029
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,6144,4096,0.05041511191262139
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,4096,0.14485066466861302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,6144,3584,0.024311110377311707
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,6144,3584,0.04621422290802002
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,3584,0.1428177754084269
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,6144,3072,0.01980177727010515
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,3584,0.013759999639458127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,10240,0.16195644272698298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,6144,3072,0.04071111149258084
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,6144,2560,0.017596445149845548
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,3072,0.14528621567620173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,3072,0.012448000411192576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,6144,2560,0.03614044520590041
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,6144,5120,0.033788445923063494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,2560,0.14473866091834173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,6144,2048,0.015257777439223396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,6144,0.01899111105336083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,6144,1536,0.012785777449607849
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,65536,0.8557368914286295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,6144,2048,0.03163911236657037
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,2048,0.010262221925788457
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,6144,1024,0.010178666975763109
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,1536,0.009319999979601966
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,6144,1536,0.02570755614174737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,4096,0.015178667174445258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,6144,1024,0.02092711130777995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,6144,768,0.008655110994974772
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,1024,0.008290666672918532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,1024,0.14543733331892225
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,6144,512,0.006948444578382704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,6144,768,0.018745778335465323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,768,0.00795288880666097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,768,0.14089511500464544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,6144,256,0.005534222142563925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,6144,512,0.017068444026841056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,512,0.007620444728268518
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,512,0.14265422026316324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,6144,128,0.004850666556093427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,6144,256,0.016122667325867545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,256,0.007265778051482306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,6144,64,0.004545777622196409
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,6144,128,0.015802666544914246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,128,0.0073315559162033936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,128,0.13430844412909612
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,6144,32,0.004559999952713649
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,6144,64,0.01534933348496755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,6144,2560,0.011415999796655444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,6144,32,0.01516711049609714
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,2048,0.14503467082977295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,5120,65536,0.27398933304680717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,5120,16384,0.07896444532606336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,1536,0.14480888843536377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,65536,0.14392977290683323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,16384,0.03905333413018121
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,5120,16384,0.1602826648288303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,5120,12288,0.06950577762391832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,16384,0.2329875628153483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,5120,10240,0.04946844445334541
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,5120,65536,0.6779004202948676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,5120,12288,0.12411377165052627
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,12288,0.030388444662094116
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,12288,0.18726310465070936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,5120,8192,0.03985688752598233
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,10240,0.02647022240691715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,5120,10240,0.1054568886756897
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,10240,0.16129333443111843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,5120,8192,0.08714844120873345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,5120,7168,0.037794666157828435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,8192,0.022280888424979314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,6144,256,0.14569955401950413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,5120,6144,0.03325422273741828
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,7168,0.1474862231148614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,7168,0.02053955528471205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,5120,5120,0.028117332193586562
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,6144,0.018950222267044913
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,5120,6144,0.06822222471237183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,6144,0.1440488894780477
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,5120,4096,0.022834666901164587
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,65536,0.8519920243157281
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,5120,0.017218665944205392
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,5120,0.14683022763994005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,5120,3584,0.02238222294383579
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,5120,4096,0.04754222101635403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,4096,0.1476960049735175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,4096,0.014726221561431885
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,5120,3072,0.01872177753183577
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,5120,3584,0.04295733239915636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,3584,0.013573333621025085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,3584,0.14185333251953125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,5120,2560,0.01719199948840671
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,5120,3072,0.03849155704180399
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,3072,0.012368000215954252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,3072,0.14810755517747667
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,5120,2560,0.03416266706254747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,5120,2048,0.015136889285511441
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,2560,0.14382044474283853
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,2560,0.011315555208259158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,5120,1536,0.012708444562223224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,5120,2048,0.02980977627966139
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,2048,0.14117688602871364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,2048,0.01015466699997584
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,5120,1024,0.009755555954244401
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,5120,1536,0.0239991115199195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,8192,0.14988444911109075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,1536,0.1481777826944987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,5120,768,0.007941332956155142
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,5120,1024,0.0199297782447603
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,5120,7168,0.07777244514889188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,1024,0.14401955074734157
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,5120,512,0.0069253332912921906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,5120,768,0.017868444323539734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,768,0.007989333735571967
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,768,0.1444853279325697
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,5120,256,0.0058133333093590206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,5120,5120,0.05654222435421414
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,5120,512,0.016196444630622864
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,512,0.007641777396202087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,512,0.1477137804031372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,5120,128,0.004960000101062986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,5120,256,0.015189333094490899
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,256,0.007265778051482306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,5120,64,0.004873777843183941
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,5120,128,0.014787554740905762
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,128,0.1336115598678589
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,128,0.007299555672539606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,5120,32,0.005590222362014983
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,5120,64,0.014560000763999091
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,5120,32,0.014471999473041959
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,4096,65536,0.24688355127970377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,4096,16384,0.07858044571346708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,1536,0.009248889154858058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,65536,0.1341786649492052
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,4096,16384,0.15084177917904323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,16384,0.03756533397568597
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,4096,12288,0.05908533599641588
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,16384,0.2308871216244168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,4096,12288,0.11481600337558323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,12288,0.02903199858135647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,12288,0.18536089526282418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,5120,256,0.14518933826022679
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,4096,10240,0.0417137775156233
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,4096,8192,0.0347671111424764
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,10240,0.025617778301239014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,4096,10240,0.09816622071795994
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,65536,0.8422009150187174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,4096,7168,0.03496444556448195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,4096,8192,0.0811742213037279
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,8192,0.02176977859603034
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,8192,0.1470915608935886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,4096,6144,0.02907466557290819
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,4096,7168,0.07238933112886217
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,7168,0.020095111595259774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,7168,0.140956441561381
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,4096,5120,0.024518221616744995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,4096,6144,0.0626595550113254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,6144,0.14628977245754665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,6144,0.018183110488785636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,4096,4096,0.02017422186003791
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,5120,1024,0.008371555142932469
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,5120,0.1402799950705634
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,4096,3584,0.01885688967174954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,4096,4096,0.04418755571047465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,4096,0.14500710699293348
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,4096,0.014183999763594734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,4096,3072,0.016717332932684157
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,4096,3584,0.04075377848413255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,3584,0.14369689093695745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,3584,0.013159111142158508
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,4096,2560,0.014856888188256158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,4096,3072,0.03570488757557339
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,3072,0.011754666765530905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,4096,2560,0.032080888748168945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,4096,2048,0.012738666600651212
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,2560,0.01089777797460556
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,2560,0.14922489060295954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,4096,2048,0.028130667077170476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,2048,0.009661333428488838
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,4096,1536,0.010312000082598792
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,4096,1536,0.02241511146227519
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,1536,0.1444986661275228
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,4096,1024,0.007864000068770515
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,1536,0.0090506664580769
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,5120,0.016581333345837064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,4096,1024,0.01889600025282966
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,4096,768,0.006606222026877933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,1024,0.008285333712895712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,1024,0.14963911639319524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,4096,768,0.01719911065366533
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,768,0.007913777397738563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,4096,512,0.005643555687533484
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,768,0.14508533477783203
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,3072,0.14081066184573704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,4096,256,0.004906666775544484
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,512,0.14732266796959773
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,10240,0.16058488686879477
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,4096,256,0.014793778459231058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,256,0.14539378219180638
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,4096,128,0.004504000147183736
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,256,0.007292444507280986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,4096,128,0.014104000396198697
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,128,0.13338222768571642
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,4096,64,0.004181333300140169
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,128,0.007286222444640265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,4096,5120,0.05349866549173991
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,4096,32,0.004193777839342753
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,4096,64,0.014087110757827759
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,4096,32,0.01408088869518704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3584,65536,0.20113778114318848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,4096,512,0.015475556254386902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,4096,512,0.0075937774446275495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,65536,0.13184266620212132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3584,16384,0.06739466720157199
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3584,65536,0.5847813288370768
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3584,16384,0.1380960014131334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,16384,0.03669333457946777
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3584,12288,0.04389333393838671
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,16384,0.22982578807406953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3584,10240,0.04221511218282911
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,12288,0.028491556644439697
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3584,12288,0.10372355249192979
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,4096,65536,0.6201439963446723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,10240,0.025063999825053748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3584,10240,0.09195822477340698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,10240,0.15883378187815347
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,65536,0.8355751037597656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3584,8192,0.03536622060669793
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3584,7168,0.0299253331290351
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,8192,0.021366222037209406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3584,8192,0.07583822144402398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3584,7168,0.067994667424096
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3584,6144,0.024534222152498033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,7168,0.14128532674577501
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,7168,0.019745777050654095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3584,5120,0.022608889473809138
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3584,6144,0.05855466922124227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,6144,0.018023111753993563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,6144,0.14715111255645752
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3584,4096,0.01965688831276364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3584,5120,0.049639112419552274
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,5120,0.016181333197487723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,5120,0.14834754996829563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,4096,2048,0.14392533567216662
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3584,3584,0.016515556308958266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3584,4096,0.04145244426197476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,4096,0.14047111405266657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,4096,0.014022222823566861
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3584,3072,0.014575110541449653
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3584,3584,0.038498666551378034
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,3584,0.012696888711717395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,12288,0.18512445025973848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3584,2560,0.012871999707486896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3584,3072,0.03404533200793796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,3072,0.14092177814907497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3584,2048,0.011382222175598145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,3072,0.011722666521867117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,2560,0.010494222243626913
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,2560,0.14519200060102674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3584,1536,0.00962577760219574
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3584,2048,0.02681066592534383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,2048,0.1441262165705363
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,2048,0.009644444617960189
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3584,1024,0.007419555551475949
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,8192,0.1472933292388916
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3584,1536,0.02152266601721446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,1536,0.1450222200817532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3584,768,0.006573333508438534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3584,1024,0.018216888109842937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,1024,0.00830666638082928
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,1024,0.14315199851989746
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3584,512,0.005537777725193236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3584,768,0.01646399994691213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,768,0.14876533879174128
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3584,256,0.004574222283230888
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3584,512,0.01536799967288971
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,512,0.007657777931955125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3584,256,0.014512888259357877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,256,0.14368800322214761
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,3584,0.1490373346540663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3584,128,0.013804444836245643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,128,0.006968888971540663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,128,0.14179466830359563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3584,64,0.0038933331767717996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3584,2560,0.03018933203485277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3584,32,0.0038888889054457345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3584,64,0.013808888693650564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3584,32,0.013760888742076026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3072,65536,0.188028441535102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,768,0.007962666451931
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,65536,0.09315022495057847
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3072,16384,0.056616001658969455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,3584,512,0.14524977737002903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3072,65536,0.5457670953538682
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3584,128,0.0041786668201287585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3072,16384,0.1318133274714152
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,16384,0.19705955187479654
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,16384,0.02544622288809882
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,12288,0.16178311241997614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3072,12288,0.09949955675337051
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3072,10240,0.03731022278467814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,12288,0.020245333512624104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3072,10240,0.08442399899164836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,10240,0.14312977261013454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,10240,0.018266666266653273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3072,8192,0.026179555389616225
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,1536,0.009009777671760982
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3072,7168,0.02544533378548092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,8192,0.1452275514602661
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3072,7168,0.06323555443021986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,7168,0.14613333013322619
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3072,6144,0.022310222188631695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,7168,0.014641778336630927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,3584,256,0.007253333098358578
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3072,6144,0.054105778535207115
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,6144,0.14117333624098036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,6144,0.013567110730542077
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3072,12288,0.04323377874162462
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3072,5120,0.04634844594531589
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,5120,0.012104888757069906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,65536,0.7023777961730957
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3072,4096,0.015944000747468736
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3072,4096,0.0389013323518965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,4096,0.15042132801479763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3072,3584,0.01462577780087789
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3072,8192,0.07164177629682753
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3072,3584,0.035916444328096174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,3584,0.14648977915445963
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3072,3072,0.013643555343151093
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,3584,0.010048000348938836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3072,3072,0.03159644537501865
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3072,2560,0.011968000067604912
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,3072,0.009369778136412302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,3072,0.14903644720713297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3072,2560,0.028090665737787884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,2560,0.14326666461096868
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3072,2048,0.010438222024175854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3072,5120,0.018657777044508193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3072,2048,0.025033776958783466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3072,1536,0.008651555412345463
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,2048,0.1461120049158732
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,5120,0.14375555515289307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3072,1024,0.006606222026877933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3072,1536,0.02058844433890449
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,1536,0.007663110891977946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,1536,0.1418186691072252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3072,768,0.005548444473081165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,4096,0.010832889212502373
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,1024,0.007303111255168915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,1024,0.1439973380830553
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3072,768,0.015840889679061044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3072,512,0.005151999907361137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,8192,0.01576622161600325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,768,0.006917333437336816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3072,256,0.0041964443193541635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3072,512,0.014700444208251106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,512,0.006893333461549547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,512,0.1432924403084649
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3072,128,0.004008888784382078
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3072,256,0.014096889230940076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,256,0.006594666590293248
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,256,0.1395502222908868
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3072,64,0.003510222252872255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3072,128,0.013530666629473368
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,3072,32,0.003842666745185852
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,128,0.1390506691402859
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,2560,0.008900444540712569
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3072,64,0.013744889034165276
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3072,32,0.013421333498424955
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2560,65536,0.16459733910030788
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,2048,0.00830666638082928
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2560,16384,0.05075022247102526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,65536,0.088100446595086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2560,16384,0.12919289535946316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,16384,0.19652622275882295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,16384,0.024671110841963027
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2560,12288,0.03433777888615926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2560,65536,0.5318062040540907
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2560,10240,0.030642665094799463
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,3072,768,0.1424942281511095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,12288,0.01973600023322635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2560,12288,0.09983822372224595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,12288,0.15983200073242188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2560,8192,0.025552888711293537
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2560,10240,0.08440088563495213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,10240,0.14047555128733316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,10240,0.017881777551439073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2560,7168,0.022312000393867493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2560,8192,0.07015733586417304
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,8192,0.14618933200836182
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,8192,0.01534400052494473
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2560,6144,0.01977422171168857
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2560,7168,0.06124799781375461
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,7168,0.014430221584108142
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,3072,128,0.006592000110281839
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,7168,0.139756441116333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2560,5120,0.017486222916179232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2560,6144,0.053024888038635254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,6144,0.1484195523791843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,6144,0.013204444613721637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2560,4096,0.01458755632241567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2560,5120,0.045824001232783
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,5120,0.14032799667782253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,5120,0.012053333222866058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2560,4096,0.03853777713245816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,4096,0.010593777729405297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,4096,0.14720355139838323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2560,3072,0.011919111013412476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2560,3584,0.03496711121665107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,3584,0.009719111025333405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,3584,0.14566666550106472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,3072,1024,0.017245334055688646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2560,2560,0.010857777463065254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2560,3072,0.03121866782506307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,3072,0.150398227903578
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,3072,0.009358222285906473
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2560,2048,0.009339555270142024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2560,2560,0.02750844425625271
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,2560,0.008615111311276754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,2560,0.14440977573394775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2560,1536,0.007602666815121968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2560,2048,0.023808000816239253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,2048,0.008344000412358178
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,2048,0.1505919959810045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,65536,0.695433775583903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2560,1024,0.0058808889653947614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,1536,0.007711110843552484
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2560,768,0.0052666668262746595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2560,1024,0.017190222938855488
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,1024,0.007317333585686154
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,1024,0.14032799667782253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2560,512,0.004864888886610667
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2560,768,0.01550222271018558
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,768,0.0070017774899800616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,768,0.14965332878960505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2560,256,0.0042035554846127825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2560,512,0.014469333820872836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2560,3584,0.013221333424250284
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,512,0.14564977751837835
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2560,128,0.0038888889054457345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2560,256,0.013787555197874704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,256,0.14078844918145075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2560,128,0.013447999954223633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,128,0.14081954956054688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,128,0.00628355559375551
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2560,32,0.0038355555799272326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2560,64,0.013101333545313941
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2560,32,0.013071999781661563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2048,65536,0.13789155748155382
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2560,1536,0.020008888509538438
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,2560,1536,0.14972355630662706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2048,16384,0.04301599992646111
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,65536,0.08410933282640244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2048,65536,0.49053064982096356
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2048,16384,0.11681866645812988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,16384,0.024524443679385718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,512,0.006621333460013072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,16384,0.1966186629401313
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2048,10240,0.025820444027582806
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,2560,256,0.006609777609507243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2048,12288,0.08816088570488824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,12288,0.15865866343180338
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2048,8192,0.021254221598307293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2560,64,0.00350222239891688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,10240,0.017768000562985737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2048,10240,0.0778684417406718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,10240,0.14129332701365152
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2048,7168,0.018199111024538677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2048,8192,0.06396177742216322
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,8192,0.0151573336786694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,8192,0.14455733034345838
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2048,6144,0.01646933290693495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2048,7168,0.05719111363093058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,7168,0.14665867222679987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2048,5120,0.014455111490355598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2048,6144,0.04922399918238322
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,6144,0.013094222380055321
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,6144,0.14734133084615073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2048,4096,0.01237511138121287
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2048,5120,0.04291911257637871
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,5120,0.011759111450778114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,5120,0.14556266201867
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2048,3584,0.011360000405046674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2048,4096,0.03592977921168009
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,4096,0.01034044474363327
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,4096,0.14526489045884874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2048,3072,0.010390222072601318
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2048,12288,0.030214223596784804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2048,3584,0.03257511059443156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,3584,0.14820178349812826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2048,2560,0.009296889106432596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2048,3072,0.029480000336964924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,3072,0.00907022257645925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,3072,0.13937777943081325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2048,2048,0.008191111187140146
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2048,2560,0.025747555825445387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,65536,0.6941288842095269
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,2560,0.1451902257071601
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2048,2048,0.022487110561794702
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2048,1536,0.0070159998204973005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,2048,0.008314666648705801
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,7168,0.01386222243309021
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2048,1536,0.019128888845443726
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2048,1024,0.00592533333433999
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,1536,0.14590044816335043
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,1536,0.007599111232492659
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2048,1024,0.016513778103722464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2048,768,0.005239999956554837
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,1024,0.14483643902672663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,1024,0.007107555866241455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2048,768,0.015063110325071545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,768,0.0069253332912921906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,768,0.1444444391462538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,3584,0.009694221946928237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2048,512,0.014089778065681458
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,512,0.14447822835710314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,12288,0.020163555939992268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2048,256,0.004222222086456087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,512,0.006631111105283101
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2048,128,0.004087999877002505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,2560,0.008662222160233391
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2048,256,0.01346666696998808
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,256,0.14478666252560085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2048,64,0.003863111138343811
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,2048,0.14717511336008707
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,128,0.006321777900060018
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2048,32,0.0038275557259718576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,2048,128,0.13774755265977648
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2048,64,0.012757333616415659
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2048,32,0.012763555679056378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1536,65536,0.10884355836444431
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1536,16384,0.0306275553173489
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,65536,0.08530488941404556
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1536,16384,0.10979021920098199
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1536,12288,0.02450399928622776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1536,65536,0.4505057864718967
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,16384,0.023901333411534626
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,2048,512,0.004558222161398994
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1536,10240,0.020419556233617995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1536,12288,0.08424088690016006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,12288,0.019215111931165058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,12288,0.15557155344221327
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1536,8192,0.01778044468826718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,65536,0.6830506854587131
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,10240,0.016921778519948322
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1536,10240,0.0712346699502733
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,10240,0.14980267153845891
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1536,7168,0.015023110641373528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,2048,256,0.006291555447710886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1536,8192,0.05793955590989855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,8192,0.014498665928840637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1536,6144,0.013618666264745923
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,8192,0.14525777763790554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1536,7168,0.05158933334880405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,7168,0.14010132683648005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1536,6144,0.04432355695300632
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,6144,0.1458337836795383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1536,5120,0.038655112187067665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1536,4096,0.010885333021481832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,5120,0.011744000017642975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,5120,0.14618933200836182
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1536,4096,0.033161779244740806
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,4096,0.01036266651418474
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1536,3584,0.009838222629494136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,4096,0.14480177561442056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,16384,0.19096177154117158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1536,3072,0.008630221916569604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,3584,0.009706666900051964
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,3584,0.14089866479237875
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1536,3072,0.02609688871436649
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1536,2560,0.007609777980380588
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,3072,0.009024889104896123
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,3072,0.1435848871866862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1536,2560,0.023704888092146978
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,2560,0.008608889248636033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,2560,0.13972889052497015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,2048,128,0.01313333296113544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,7168,0.013239999612172445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1536,2048,0.021363554729355708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,2048,0.13934755325317383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1536,1536,0.005995555470387141
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,2048,0.008273777862389883
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,6144,0.012424888710180918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1536,1024,0.004868444469239977
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1536,1536,0.017880888448821176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,1536,0.007666666474607255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,1536,0.14014932844373915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1536,768,0.004536889079544279
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1536,1024,0.015607999430762397
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,1024,0.1398142178853353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1536,512,0.0041680000722408295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1536,768,0.014468444718254937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,768,0.0069155556460221606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,768,0.14543644587198892
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1536,256,0.0035182221068276297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1536,512,0.013447999954223633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,512,0.006618666566080517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,512,0.14034844769371882
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1536,3584,0.02983199887805515
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1536,128,0.0031795555518733132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,256,0.006264888991912206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1536,128,0.0127368892232577
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,128,0.13394310739305285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1536,64,0.0031475555151700974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1536,2048,0.006650666809744305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1536,32,0.0032746667663256326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,128,0.006315555423498154
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1536,5120,0.01274577776590983
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1024,65536,0.08034666379292806
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,1536,1024,0.007232888705200619
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1024,65536,0.44192179044087726
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,65536,0.08378666639328003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1024,16384,0.025605332520272996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1536,256,0.01312355531586541
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,1536,256,0.14008977678087023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,65536,0.6756426493326823
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1024,16384,0.10860533184475368
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,16384,0.1916053295135498
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1024,12288,0.0203448881705602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,16384,0.02270222206910451
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1024,10240,0.01794666714138455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,12288,0.01826400061448415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1024,12288,0.08208355638715956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,12288,0.154639999071757
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1024,10240,0.06947377655241224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,10240,0.14147999551561144
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1024,8192,0.015562666787041558
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1024,7168,0.013797333670987023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,10240,0.01651822196112739
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1536,64,0.012412444584899478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1536,32,0.012149333126015134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,8192,0.014087110757827759
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1024,6144,0.012531555361217923
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,7168,0.01312711089849472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1024,7168,0.049901333120134145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,7168,0.15101422203911677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1024,5120,0.011039999624093374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1024,6144,0.04409866531689962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,6144,0.012163555456532372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,6144,0.14525866508483887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1024,5120,0.03817244370778402
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1024,4096,0.008997333546479544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,5120,0.011358222199810876
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1024,3584,0.008265777594513362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1024,4096,0.03267733256022135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,4096,0.15038932694329157
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1024,3072,0.007625777688291337
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1024,3584,0.028873778051800195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,3584,0.009452444811662039
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,3584,0.14745244714948866
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1024,2560,0.006813333266311222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1024,3072,0.026064889298544988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,3072,0.008997333546479544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,3072,0.14042933781941733
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1024,2048,0.005951111101441913
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1024,2560,0.023538667294714186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,2560,0.008608000146018134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,2560,0.150416890780131
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1024,1536,0.005234666582610872
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1024,2048,0.020968000094095867
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,2048,0.007978666987684038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,2048,0.1480462286207411
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1024,1024,0.004515555583768421
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,8192,0.14485155211554632
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1024,1536,0.017864889568752713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,1536,0.14036799801720515
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,1536,0.007635555333561367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1024,768,0.0041599998043643106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1024,1024,0.015460444821251763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,1024,0.0069617778062820435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,1024,0.14458133114708796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1024,512,0.003882666842805015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1024,768,0.014123555686738757
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,768,0.13924354977077907
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,768,0.0068400001360310455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1024,256,0.0034862222770849862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1024,512,0.013432000246312885
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,512,0.1483102242151896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1024,256,0.012740444805887012
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,256,0.006292444550328785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1024,128,0.012409778104888068
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,128,0.13670489523145887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1024,64,0.0035111109415690103
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,128,0.006252444452709622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1024,32,0.0032257777121331957
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1024,64,0.012107555237081317
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1024,32,0.012056888805495368
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,768,65536,0.07309155331717597
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,1024,8192,0.05646489063898722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,65536,0.0827253328429328
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,768,16384,0.0207324441936281
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,5120,0.1457644436094496
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,512,0.006633777585294511
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,1024,128,0.0034133332471052804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,1024,256,0.14139910538991293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,768,12288,0.01963200006220076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,16384,0.19205244382222494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,65536,0.6747306717766656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,12288,0.15431643856896296
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,12288,0.0184888889392217
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,768,10240,0.01683911184469859
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,768,10240,0.06510044468773736
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,10240,0.1494364473554823
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,768,8192,0.014745778507656522
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,10240,0.016419554750124615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,768,8192,0.051058666573630444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,8192,0.1427324480480618
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,768,7168,0.013261333107948303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,8192,0.014058666096793281
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,768,65536,0.4022124343448215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,768,7168,0.04676444331804911
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,1024,4096,0.010055999788973067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,7168,0.15080355273352727
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,768,6144,0.01163200040658315
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,7168,0.013000888956917657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,768,16384,0.09859466552734375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,768,5120,0.010032888915803697
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,16384,0.02272622287273407
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,6144,0.012097777591811286
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,6144,0.14819199509090847
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,768,4096,0.008637333081828224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,768,5120,0.033964445193608604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,768,12288,0.07480533255471124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,5120,0.011392888923486074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,5120,0.14843111568027073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,768,3584,0.008217777642938826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,768,4096,0.029948443174362183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,4096,0.14317688677046034
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,4096,0.010045333041085137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,768,3072,0.007296000089910295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,768,3584,0.027003554834259882
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,3584,0.009364444348547194
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,768,2560,0.006624889042642381
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,768,3072,0.0243520008193122
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,3072,0.1445715559853448
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,768,2560,0.022279999322361414
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,2560,0.008512000242869059
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,768,2048,0.019720888800091214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,2048,0.14320977528889975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,768,1536,0.005214222189452913
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,2048,0.007992888490358988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,768,1536,0.016945777667893302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,1536,0.13904799355400935
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,768,1024,0.004510222209824456
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,1536,0.0075635554061995606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,768,6144,0.03964711229006449
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,768,1024,0.014833778142929077
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,1024,0.006953777538405524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,1024,0.1446897718641493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,3584,0.1473368936114841
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,768,512,0.0038524443904558816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,768,0.006804444309737947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,768,0.14461244477166071
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,768,2048,0.0058293334311909145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,768,256,0.0034826666944556763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,512,0.006641777853171031
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,512,0.14555999967787
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,768,256,0.0124195557501581
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,256,0.0063493334584765965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,768,128,0.0031431110368834603
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,768,128,0.012145777543385824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,128,0.1373680035273234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,768,64,0.002971555623743269
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,128,0.006256888724035687
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,768,32,0.002991111121243901
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,768,64,0.012075555821259817
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,512,65536,0.04560000035497877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,768,768,0.003946666502290302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,768,768,0.013800000150998434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,768,3072,0.008918222453859117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,512,16384,0.019291554888089497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,65536,0.08224888642628987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,768,512,0.01308444473478529
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,2560,0.14368177784813776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,512,65536,0.40099912219577366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,512,12288,0.01610844499535031
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,16384,0.02232711182700263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,512,16384,0.09746311108271281
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,16384,0.18844800525241426
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,512,10240,0.014925332532988654
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,65536,0.6649129125807021
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,12288,0.15513511498769125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,12288,0.01755555636352963
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,512,8192,0.013551111022631327
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,512,10240,0.06264622343911065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,10240,0.015855110353893705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,512,7168,0.012525333298577202
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,512,8192,0.049350221951802574
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,8192,0.14374666743808323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,8192,0.013779555757840475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,512,6144,0.011012444065676795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,512,7168,0.04378666811519199
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,7168,0.012821333275900947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,7168,0.147523562113444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,768,32,0.011782222323947482
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,512,5120,0.009413333402739631
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,512,6144,0.03963377740648057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,6144,0.01203644441233741
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,6144,0.1403226719962226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,512,4096,0.008267555799749162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,512,5120,0.03437688946723938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,5120,0.14906844827863905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,5120,0.01128799964984258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,512,3584,0.007343110938866933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,512,4096,0.02913244565327962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,4096,0.010010666317409938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,512,3584,0.026722666290071275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,3584,0.009311111436949836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,512,3072,0.007233777807818518
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,768,256,0.14386044608222112
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,512,12288,0.07401155763202243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,512,3072,0.02370400064521366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,3072,0.008741333252853816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,3072,0.14349867237938776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,10240,0.14971021811167398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,512,2048,0.005595555735958948
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,2560,0.008344000412358178
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,2560,0.1436648898654514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,512,2048,0.019534221953815885
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,2048,0.007937777373525832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,2048,0.1450951099395752
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,512,1536,0.0052328887912962176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,1536,0.007601777712504069
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,512,1536,0.01684888866212633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,1536,0.1453440056906806
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,512,1024,0.0042151109211974675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,512,1024,0.014659555421935188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,1024,0.0070257774657673305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,1024,0.14231733481089273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,512,768,0.004197333421972063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,512,768,0.013428444663683573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,768,0.006871999965773688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,768,0.14388533433278403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,512,512,0.0038657776183552216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,512,512,0.0127368892232577
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,512,2560,0.006280888699822956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,512,256,0.0032142222755485107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,512,2560,0.02164266672399309
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,512,256,0.012373333175977072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,256,0.14323111375172934
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,512,128,0.0031884445084465873
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,256,0.006240889016124938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,512,128,0.011882666912343768
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,512,64,0.0031724443866146933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,128,0.006280000011126201
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,512,32,0.0028515555378463534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,512,64,0.011696889168686338
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,256,65536,0.03155111273129781
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,512,32,0.011756444142924415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,4096,0.14586755964491102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,256,16384,0.013770666387346057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,65536,0.08063733577728271
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,3584,0.1414533323711819
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,256,65536,0.4004133277469211
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,256,16384,0.09513866901397705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,512,512,0.006607111129495833
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,256,12288,0.01148000028398302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,16384,0.1899706655078464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,16384,0.021704889006084863
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,512,0.1423742241329617
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,256,10240,0.010459555519951714
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,65536,0.6686453289455838
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,12288,0.01762488815519545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,256,12288,0.07097333007388644
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,12288,0.15536444716983372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,256,8192,0.009653333160612319
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,10240,0.015819556183285184
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,256,7168,0.009296000003814697
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,8192,0.013791999883121915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,256,8192,0.049345778094397656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,8192,0.1456471151775784
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,256,6144,0.008325333396593729
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,512,128,0.13821510473887125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,7168,0.14533689286973742
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,256,5120,0.008237332933478886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,7168,0.012808000048001608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,256,6144,0.038570665650897555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,6144,0.01202133297920227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,256,4096,0.007629333270920648
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,256,5120,0.03310044606526693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,5120,0.15114311377207437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,256,4096,0.02843377656406826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,4096,0.14416889349619547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,256,3584,0.007262222468852997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,256,3584,0.02640266716480255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,3584,0.14853510591718885
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,256,3072,0.00656444455186526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,3584,0.009329777624871995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,256,3072,0.023644444015291
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,3072,0.14458489418029785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,256,2560,0.006227555374304454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,3072,0.00867022242810991
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,256,7168,0.044405334525638156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,256,2560,0.0217848883734809
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,2560,0.008286221987671321
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,2560,0.1454755597644382
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,6144,0.1404391129811605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,5120,0.011181332998805575
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,256,2048,0.0192586663696501
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,2048,0.14537599351671007
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,256,1536,0.004906666775544484
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,2048,0.007999999655617608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,4096,0.009928888744778102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,256,1024,0.0041759999261962045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,256,1536,0.01682133310370975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,1536,0.14780444569057888
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,1536,0.007487110793590546
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,256,768,0.003863999827040566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,256,1024,0.01445777714252472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,1024,0.0069724445541699724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,1024,0.14912977483537462
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,256,512,0.003674666500753827
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,256,768,0.013472889032628802
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,768,0.006629333313968446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,768,0.14955200089348689
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,256,256,0.0031804444475306403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,256,10240,0.06172177526685926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,256,512,0.012771555946932899
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,512,0.006280000011126201
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,512,0.145304004351298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,256,128,0.0030257778449190986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,256,256,0.012075555821259817
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,256,0.00629066675901413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,256,0.139756441116333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,256,64,0.002817777709828483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,256,128,0.012078222301271228
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,256,32,0.002875555513633622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,256,128,0.00628355559375551
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,128,0.13840177324083117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,256,2048,0.0052622221410274506
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,128,65536,0.026320889592170715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,256,64,0.011699555648697747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,128,16384,0.012462221913867526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,256,32,0.011722666521867117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,65536,0.07966933647791545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,128,16384,0.09603910975986057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,16384,0.021514667405022517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,128,12288,0.010378667049937779
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,16384,0.19078132841322157
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,128,10240,0.009641778137948778
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,128,65536,0.39966310395134824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,12288,0.01755911111831665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,128,12288,0.07266400257746379
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,128,8192,0.008624888956546783
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,10240,0.14853689405653211
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,65536,0.6685306760999891
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,128,7168,0.00828888929552502
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,128,8192,0.04924355612860786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,8192,0.1415439976586236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,256,10240,0.1434719959894816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,128,7168,0.043176889419555664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,7168,0.14143910672929552
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,128,6144,0.00794311116139094
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,7168,0.01276000009642707
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,128,5120,0.007306666837798224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,6144,0.011972444752852121
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,6144,0.14046843846638998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,128,5120,0.03313510947757297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,128,4096,0.007634667058785756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,5120,0.011056889262464313
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,5120,0.1435226731830173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,128,4096,0.02846933404604594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,4096,0.009967111051082611
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,4096,0.14044000042809382
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,128,3584,0.007253333098358578
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,3584,0.009295110901196798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,3584,0.1400675508711073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,12288,0.1539413399166531
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,128,3072,0.006559111177921295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,128,10240,0.06117866436640421
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,10240,0.015822221835454304
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,128,2560,0.005616888817813661
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,128,3072,0.023864888482623633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,3072,0.00869066682126787
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,3072,0.14437599976857504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,128,2048,0.0051857779423395796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,8192,0.013718222578366598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,128,2560,0.02166666587193807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,2560,0.00831200016869439
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,2560,0.1478195587793986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,128,1536,0.00452888881166776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,128,2048,0.019288000133302476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,2048,0.00795199970404307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,2048,0.14076088534461126
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,1536,0.007328000333574083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,128,1536,0.01680355601840549
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,1536,0.14665422174665663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,128,768,0.00350577798154619
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,128,6144,0.03749600052833557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,128,1024,0.014449778530332776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,1024,0.1413608921898736
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,128,512,0.003301333429084884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,128,768,0.013473778135246701
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,768,0.006583111153708563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,768,0.1415662235683865
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,128,256,0.0031431110368834603
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,128,512,0.012720888687504662
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,512,0.006285333385070165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,512,0.1509688827726576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,128,128,0.002830222249031067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,128,256,0.012082666986518435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,256,0.14537421862284342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,128,128,0.011740444435013665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,128,64,0.002830222249031067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,128,0.006275555739800136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,128,32,0.002532444480392668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,128,64,0.011388444238238864
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,128,32,0.011702222128709158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,64,65536,0.02572000026702881
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,64,16384,0.009570666485362584
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,128,1024,0.004210666649871402
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,64,12288,0.008485332959228093
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,64,16384,0.09587466716766357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,64,10240,0.009018667042255402
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,64,65536,0.39906220965915257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,1024,0.0069724445541699724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,64,8192,0.00833511104186376
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,64,12288,0.07252177927229139
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,64,7168,0.008269333177142674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,64,10240,0.06074755721622043
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,64,6144,0.007326222128338284
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,64,5120,0.008119111259778341
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,64,8192,0.04822044571240743
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,64,7168,0.04314844475852119
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,64,6144,0.03760711020893521
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,768,128,256,0.00629688882165485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,64,5120,0.03254488772816128
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,64,3072,0.007038222418891058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,64,4096,0.028117332193586562
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,128,3584,0.026406221919589575
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,64,2560,0.006608888920810487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,64,3584,0.026100445124838088
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,64,2048,0.005195555587609609
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,64,3072,0.023362666368484497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,64,1536,0.004552888787455029
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,64,1024,0.0038559999730851916
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,64,2560,0.02132266594303979
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,64,2048,0.0192275560564465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,64,1536,0.01647555496957567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,64,1024,0.014584889014561972
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,64,512,0.0032071111102898917
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,64,256,0.003034666594531801
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,64,768,0.013418667018413544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,64,128,0.0028577778074476454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,64,512,0.012883555557992725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,64,64,0.0025315555847353404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,64,256,0.011762667033407422
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,64,32,0.002789333255754577
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,64,128,0.011719110939237805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,64,64,0.011730666789743634
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,32,65536,0.019813333948453266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,64,32,0.011685333318180509
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,32,16384,0.014152889450391134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,32,12288,0.009508444203270806
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,64,4096,0.007620444728268518
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,32,10240,0.008716444174448649
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,64,3584,0.007302222152551015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,32,12288,0.07232533560858832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,32,8192,0.008354666332403818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,32,7168,0.007958222594526079
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,32,10240,0.0610719985432095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,32,6144,0.007642666498819987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,32,7168,0.04227822356753879
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,32,5120,0.007291555404663086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,32,6144,0.03730666637420654
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,32,4096,0.0075866662793689305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,32,3584,0.0075786668393347
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,32,5120,0.03259288933542039
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,32,4096,0.028195556667115953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,32,3072,0.006948444578382704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,32,2560,0.006653333289755716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,32,3584,0.026130666335423786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,32,2048,0.005254222287072076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,32,3072,0.02347466680738661
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,32,2560,0.02144977781507704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,32,2048,0.019254222512245178
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,32,1024,0.0038844446341196695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,32,1536,0.01640266676743825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,32,768,0.003544888976547453
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,32,16384,0.09584177864922418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,32,1024,0.014467555615637036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,32,512,0.0032026666320032547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,32,256,0.002887111157178879
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,32,65536,0.3994293212890625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,32,512,0.012493333054913415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,32,128,0.0028497777465316984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,32,8192,0.048287110196219556
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,32,256,0.012039110892348819
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,32,32,0.002550222186578645
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,32,128,0.011687999798191918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,64,768,0.0034808889031410217
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,32,64,0.0118568887313207
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,32,32,0.011428444749779172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,32,1536,0.00453599997692638
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,65536,16384,0.5951929092407227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,16384,0.24725243780348036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,65536,16384,0.7119404474894205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,768,32,768,0.013101333545313941
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,65536,12288,0.4567555321587457
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,768,32,64,0.0025066667132907445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,65536,12288,0.5453911357455784
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,65536,10240,0.37799912028842503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,12288,0.18328621652391222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,65536,10240,0.4385040071275499
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,65536,8192,0.31852176454332143
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,10240,0.1528302166197035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,65536,8192,0.3498728805118137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,16384,0.6977981991238065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,12288,0.5109608968098959
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,8192,0.1270844406551785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,10240,0.42609866460164386
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,65536,7168,0.3454337649875217
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,7168,0.11273422506120469
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,8192,0.35285245047675234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,65536,6144,0.24877601199679902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,65536,6144,0.2749271127912733
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,65536,5120,0.19842755794525146
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,6144,0.09806488619910346
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,7168,0.31923643747965497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,65536,5120,0.22663643625047472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,65536,4096,0.16306222809685603
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,65536,7168,0.2625519964430067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,5120,0.08479822344250149
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,6144,0.2802471054924859
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,65536,4096,0.18885955545637342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,4096,0.06923644410239325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,65536,3584,0.16182666354709202
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,5120,0.24737599160936144
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,65536,3072,0.12317244211832683
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,4096,0.20222311549716523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,3584,0.1852124399609036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,65536,3072,0.14140355587005615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,65536,2560,0.1090008881356981
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,3072,0.054213331805335156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,65536,2048,0.08463022443983291
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,3072,0.16637421978844538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,65536,2560,0.12424800131056045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,2560,0.04783022072580126
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,2560,0.15011466874016657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,65536,1536,0.07745599746704102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,65536,2048,0.10226577520370483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,2048,0.04118844535615709
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,2048,0.14266933335198298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,65536,1024,0.050042665667004056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,65536,1536,0.08246222469541761
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,1536,0.03347644540998671
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,1536,0.14740978346930608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,65536,768,0.03564000129699707
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,65536,3584,0.13438577122158474
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,1024,0.1460497776667277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,1024,0.026695999834272597
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,65536,1024,0.0647031135029263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,65536,512,0.030018667380015056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,65536,768,0.05424355467160543
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,768,0.15237955252329508
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,65536,256,0.020080000162124634
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,512,0.14094221591949463
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,65536,512,0.04741511080000135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,512,0.021499555971887376
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,65536,128,0.019332443674405415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,256,0.14714044994778103
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,65536,256,0.042380445533328585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,256,0.0196079992585712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,65536,128,0.03975377811325921
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,65536,128,0.13563556141323513
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,128,0.019352888067563374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,65536,32,0.016971555021074083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,65536,64,0.039156443542904325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,65536,32,0.03910311063130697
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,16384,65536,0.6045226520962185
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,768,0.023856888214747112
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,16384,16384,0.16939732763502333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,65536,3584,0.06144977940453423
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,16384,65536,0.8866000175476074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,16384,0.0694613324271308
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,16384,12288,0.13420355319976807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,16384,16384,0.21930132971869576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,16384,12288,0.167085329691569
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,12288,0.05298133360015022
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,16384,10240,0.10375022225909764
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,16384,0.27287912368774414
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,12288,0.21767377853393555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,16384,10240,0.14114756054348415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,16384,8192,0.09353688690397476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,10240,0.045774221420288086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,10240,0.1846924490398831
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,16384,8192,0.11463377210828994
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,65536,0.2656213442484538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,16384,7168,0.07533599932988484
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,8192,0.0377422206931644
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,16384,6144,0.0777733325958252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,16384,7168,0.10288177596198188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,7168,0.03369333346684774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,65536,64,0.01649777756796943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,7168,0.14639555083380804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,16384,5120,0.0642168919245402
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,16384,6144,0.0912097758717007
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,6144,0.02973777718014187
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,65536,1.0073013305664062
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,6144,0.14769244194030762
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,16384,5120,0.07811644342210558
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,5120,0.026125333375400964
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,5120,0.15160266558329263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,16384,3584,0.04447911183039347
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,16384,4096,0.06386488676071167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,4096,0.14449689123365614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,4096,0.022104889154434204
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,16384,3072,0.038418667183982
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,16384,3584,0.05847822295294868
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,3584,0.14625599649217394
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,3584,0.020004444652133517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,16384,2560,0.03191555539766947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,16384,3072,0.0521982212861379
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,3072,0.15115733941396078
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,3072,0.01815111107296414
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,16384,2048,0.02592977715863122
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,8192,0.15825243790944418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,2560,0.14777599440680608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,2560,0.01684888866212633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,16384,1536,0.021498666869269475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,16384,2048,0.037552889850404524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,2048,0.014745778507656522
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,16384,1536,0.03149688906139798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,16384,1024,0.015646222564909194
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,16384,4096,0.04938755432764689
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,1024,0.1432853274875217
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,1024,0.010910222099887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,16384,768,0.01310755560795466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,16384,2560,0.0454364452097151
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,16384,768,0.023059555225902136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,768,0.010011555420027839
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,768,0.1453582180870904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,16384,512,0.010148444109492833
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,1536,0.012816888590653738
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,16384,512,0.0207360006040997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,512,0.14109688334994847
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,16384,256,0.007631111476156447
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,512,0.009294222626421187
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,16384,1024,0.026001777913835313
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,16384,128,0.006628444625271692
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,16384,256,0.018869333797030978
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,256,0.14387910895877415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,256,0.008692444198661381
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,16384,64,0.005872000008821487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,16384,128,0.018188445104493033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,16384,32,0.005888000130653381
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,128,0.13896800412072077
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,16384,128,0.008965333302815756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,16384,64,0.018030222919252183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,16384,32,0.017881777551439073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,12288,16384,0.12601510683695474
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,12288,65536,0.4946346812778049
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,12288,16384,0.18288532892862955
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,2048,0.14685155285729304
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,65536,0.2308382193247477
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,12288,12288,0.1035262213812934
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,16384,0.05826489130655924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,12288,12288,0.14125688870747885
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,12288,10240,0.08848977751202053
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,16384,0.24158755938212076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,12288,0.045012444257736206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,12288,0.18607021702660453
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,10240,0.038986666334999934
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,12288,10240,0.11947555012173122
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,12288,8192,0.07759644587834676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,10240,0.17061155372195771
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,12288,8192,0.09698488977220322
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,8192,0.030080888006422255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,8192,0.14465778403811985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,12288,7168,0.08612533410390218
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,7168,0.02894044584698147
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,12288,6144,0.05344533258014255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,65536,0.8340053558349609
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,16384,1536,0.14596888754102919
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,12288,6144,0.0765226682027181
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,6144,0.15203732914394802
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,6144,0.02584533393383026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,12288,5120,0.06710488928688897
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,12288,4096,0.04145777887768216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,5120,0.1499511135949029
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,12288,4096,0.05465777715047201
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,4096,0.15086755487653944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,12288,3584,0.034428444173600935
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,4096,0.019550222489568923
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,12288,7168,0.06911555263731214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,12288,3584,0.04814133379194471
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,3584,0.1469564437866211
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,12288,3072,0.03199111090766059
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,7168,0.14835111300150552
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,12288,3072,0.043508443567487926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,12288,65536,0.7405653529696994
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,3072,0.15064088503519693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,12288,2560,0.026985777748955622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,3072,0.015340444114473132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,12288,5120,0.04920622375276354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,12288,2048,0.022443556123309668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,12288,2560,0.039637333816952176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,2560,0.014780445231331719
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,2560,0.15145244863298205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,12288,2048,0.03228800164328681
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,12288,1536,0.016688888271649677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,2048,0.1473697821299235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,2048,0.013111111190583972
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,12288,1024,0.013449778159459433
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,12288,1536,0.02773422168360816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,1536,0.010847999817795224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,1536,0.15136533313327366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,12288,768,0.011514667007658215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,12288,1024,0.023248000277413264
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,1024,0.1449671056535509
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,12288,768,0.020237333244747586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,12288,512,0.009679111341635386
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,768,0.14502755800882974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,768,0.008661333057615492
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,512,0.00831288927131229
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,3584,0.017253332667880587
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,12288,256,0.007269333634111617
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,12288,256,0.017198221551047433
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,256,0.145541336801317
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,12288,128,0.006730666591061487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,256,0.007941332956155142
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,12288,128,0.016798221402698092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,128,0.13516088326772055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,5120,0.02237155536810557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,128,0.00795999997191959
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,12288,32,0.006579555571079254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,12288,64,0.01648177703221639
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,12288,32,0.016492444607946608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,12288,1024,0.009304888546466827
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,12288,512,0.018223111828168232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,10240,65536,0.40499644809299046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,12288,512,0.14469156000349256
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,10240,16384,0.11712443828582764
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,768,128,128,0.13726133770412868
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,10240,16384,0.16079910596211752
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,10240,65536,0.6498995357089573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,10240,12288,0.09217955668767293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,65536,0.18585422303941515
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,12288,64,0.006211555666393704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,12288,0.03959733247756958
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,16384,0.23310666614108613
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,10240,10240,0.0821360018518236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,12288,0.18462844689687094
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,10240,8192,0.06631822056240506
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,10240,10240,0.10625511407852173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,10240,0.0347324444188012
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,10240,0.16020622518327501
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,10240,7168,0.05682933330535889
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,10240,8192,0.08678311109542847
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,8192,0.028324445088704426
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,8192,0.14390667279561362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,10240,6144,0.038916445440716214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,7168,0.02574666672282749
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,10240,7168,0.0773493316438463
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,7168,0.14131911595662436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,10240,6144,0.0681155522664388
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,10240,5120,0.04135200050142076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,65536,0.8339369032118055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,6144,0.02308533257908291
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,6144,0.14785422219170466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,10240,4096,0.032540443870756365
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,10240,5120,0.05956266985999214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,5120,0.14370755354563394
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,10240,3584,0.02895555562443203
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,10240,4096,0.048800001541773476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,16384,0.0500257776843177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,4096,0.14721866448720297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,4096,0.018183110488785636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,10240,3072,0.02458933326933119
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,10240,3584,0.043256001340018384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,3584,0.016336000627941556
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,3584,0.1424702273474799
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,10240,12288,0.12409510877397324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,10240,2560,0.02201955517133077
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,10240,3072,0.039304888910717435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,3072,0.015048889650238885
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,2560,0.014009777042600842
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,2560,0.13998844888475206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,10240,1536,0.013717333475748697
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,10240,2048,0.02944533361328973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,2048,0.1407973368962606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,2048,0.012286221815480126
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,10240,1024,0.0107022225856781
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,10240,1536,0.02545333405335744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,1536,0.01035733355416192
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,1536,0.14454132980770534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,10240,768,0.009294222626421187
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,10240,1024,0.021346666746669348
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,1024,0.008976000050703684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,10240,768,0.0188755558596717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,768,0.008380444513426887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,10240,512,0.007267555428875818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,5120,0.020409777760505676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,10240,512,0.017125333348910015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,512,0.14694400628407797
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,10240,256,0.006258666515350342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,512,0.008164444731341468
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,10240,256,0.016202666693263583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,256,0.14033867253197566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,10240,128,0.00552977787123786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,10240,2048,0.017484444710943434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,3072,0.1465582185321384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,256,0.007589333587222629
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,10240,2560,0.03532177872127957
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,10240,64,0.005229333208666907
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,10240,32,0.005199999858935674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,10240,128,0.015828443898095023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,10240,128,0.007607999775144789
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,10240,64,0.01552177800072564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,10240,32,0.015527110960748462
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,8192,65536,0.2557306554582384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,1024,0.1409777800242106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,8192,16384,0.08061333497365315
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,768,0.14572800530327693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,65536,0.14449244075351292
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,8192,65536,0.5672604242960612
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,8192,12288,0.07903822263081868
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,16384,0.03921511107020908
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,16384,0.19462045033772787
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,8192,12288,0.10848977830674912
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,8192,10240,0.05684533384111193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,12288,0.1591528919008043
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,8192,10240,0.0926355587111579
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,10240,128,0.13506133026546902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,10240,0.1418586704466078
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,8192,8192,0.037787556648254395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,8192,8192,0.07519644498825073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,8192,0.022043555974960327
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,8192,0.13935821586185032
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,65536,0.6862053341335721
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,8192,7168,0.04473955432573954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,8192,6144,0.03870222303602431
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,8192,7168,0.06721066766315036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,7168,0.020651555723614164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,8192,6144,0.05793599949942695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,8192,16384,0.14271556006537542
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,6144,0.018572444717089336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,6144,0.14763466517130533
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,8192,4096,0.025747555825445387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,8192,5120,0.05052177773581611
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,5120,0.14140355587005615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,5120,0.01666755477587382
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,12288,0.030794666873084173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,8192,3584,0.023285332653257582
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,8192,4096,0.04230133361286587
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,4096,0.014945778581831189
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,4096,0.1427475611368815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,8192,3072,0.020052444603708055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,8192,3584,0.03865333398183187
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,3584,0.14526489045884874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,8192,2560,0.01722400055991279
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,8192,3072,0.03509955604871114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,3072,0.01239377756913503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,3072,0.14120977454715306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,8192,2048,0.014354666074117025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,8192,2560,0.032079110542933144
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,2560,0.011439999772442712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,2560,0.14449955357445612
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,8192,1536,0.012055111428101858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,8192,2048,0.026820444398456152
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,2048,0.14504710833231607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,2048,0.009674666656388177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,8192,1024,0.010015111002657149
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,8192,1536,0.023078221413824294
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,1536,0.13963199986351862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,8192,5120,0.0306311117278205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,8192,1024,0.01907466683122847
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,1024,0.008046222229798635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,8192,768,0.01754755609565311
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,768,0.14053422874874538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,8192,512,0.006903999795516332
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,768,0.007637333538797166
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,8192,512,0.016171556380059984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,3584,0.013555555707878537
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,8192,256,0.005524444497293896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,512,0.145541336801317
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,512,0.00737955586777793
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,256,0.14659289518992105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,8192,128,0.004875555634498596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,8192,256,0.015496888094478183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,256,0.0069520001610120135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,8192,128,0.015145777000321282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,7168,0.14099021752675375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,128,0.13895022869110107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,8192,64,0.004184000194072723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,8192,32,0.004854222138722737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,1536,0.008998221821255153
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,8192,768,0.008316444853941599
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,8192,32,0.014841778410805596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,8192,1024,0.14496000607808432
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,7168,16384,0.09084088934792413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,7168,65536,0.30795110596550834
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,65536,0.13397599591149226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,7168,16384,0.13708089457617864
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,7168,65536,0.5501911375257704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,16384,0.19361510541703966
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,16384,0.038310223155551486
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,7168,12288,0.10566400157080756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,12288,0.15875021616617838
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,12288,0.030591110388437908
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,7168,10240,0.05725955300860935
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,8192,64,0.015108444624476962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,7168,10240,0.09028444687525432
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,10240,0.14611289236280653
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,7168,8192,0.045315557056003146
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,10240,0.02608266638384925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,7168,8192,0.07455466853247748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,8192,0.021913778450753953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,8192,0.14320354991488987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,7168,7168,0.040390223264694214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,10240,0.02644799980852339
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,7168,7168,0.06672888994216919
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,7168,0.01999022232161628
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,7168,0.14186399512820774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,7168,6144,0.03654399845335219
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,7168,5120,0.029655112160576716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,6144,0.018588445252842374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,7168,6144,0.05771733654869927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,6144,0.14578666951921251
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,7168,5120,0.049138665199279785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,7168,4096,0.0244168887535731
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,5120,0.14423022005293104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,5120,0.01651822196112739
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,8192,128,0.006962666908899943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,7168,3584,0.0229413327243593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,7168,4096,0.04147555430730184
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,4096,0.14650400479634604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,4096,0.014492443866199918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,7168,3072,0.020101333657900494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,7168,3584,0.037979556454552546
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,3584,0.14132889111836752
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,65536,0.6815395355224609
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,7168,3072,0.03468888998031616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,3072,0.14620799488491482
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,3072,0.01237422227859497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,7168,2048,0.015094222293959724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,2560,0.14451911714341906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,7168,2560,0.030942221482594807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,7168,2048,0.02608355548646715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,7168,1536,0.012655110822783576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,2048,0.009680888719028896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,7168,12288,0.06881778107749091
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,7168,1024,0.01015733347998725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,7168,1536,0.022616000639067754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,1536,0.008766222331258986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,1536,0.15168889363606772
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,7168,768,0.008625778059164682
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,7168,1024,0.018572444717089336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,1024,0.007983110845088959
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,1024,0.14131555292341444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,7168,512,0.007194666398896112
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,7168,768,0.016865778300497267
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,768,0.007651555869314406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,768,0.14484443929460314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,7168,512,0.015849777393870883
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,7168,256,0.005564444594913059
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,512,0.14929244253370497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,512,0.007261333366235097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,3584,0.013439999686347114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,7168,128,0.005120888766315249
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,256,0.14605511559380427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,256,0.006958222223652734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,7168,64,0.0046746668716271715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,7168,128,0.014478221535682678
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,128,0.006933333145247565
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,128,0.13478044668833414
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,7168,32,0.006288888967699475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,7168,2560,0.011362666885058085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,7168,64,0.014210666219393412
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,7168,32,0.014468444718254937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,7168,2048,0.14423733287387425
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,6144,16384,0.07454933060540093
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,6144,65536,0.29270132382710773
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,65536,0.12833955552842882
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,16384,0.03726844324005975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,6144,16384,0.12712177965376112
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,6144,12288,0.05182222194141812
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,6144,65536,0.5046124458312988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,16384,0.1925875610775418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,6144,10240,0.046428445312711925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,12288,0.029416001505321924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,12288,0.15808088249630398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,6144,10240,0.08446488777796428
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,6144,8192,0.04066933194796244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,10240,0.14763555261823866
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,7168,256,0.015150222513410779
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,7168,2560,0.017864889568752713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,6144,7168,0.034550223085615374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,6144,8192,0.06886400116814508
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,8192,0.02132355504565769
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,8192,0.1448462274339464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,6144,6144,0.030064000023735896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,6144,7168,0.06165511078304715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,65536,0.6715884208679199
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,7168,0.14298133055369058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,6144,6144,0.053342223167419434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,6144,5120,0.025938666529125635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,6144,0.14770221710205078
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,6144,0.017989334132936265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,6144,4096,0.021295999487241108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,5120,0.1457200050354004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,5120,0.016214221715927124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,6144,4096,0.039057777987586126
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,4096,0.01423733267519209
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,4096,0.14416000578138563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,6144,3584,0.03558311197492812
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,3584,0.14578133159213597
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,6144,12288,0.10170755783716838
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,6144,3072,0.01665511065059238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,3584,0.013015111287434896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,10240,0.025421332981851365
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,6144,3072,0.03257955445183648
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,6144,2560,0.01514400045077006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,3072,0.14491377936469182
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,3072,0.011885333392355176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,6144,2048,0.013050666285885705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,6144,2560,0.029603554142846003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,2560,0.011028444601429833
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,2560,0.15120089054107666
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,6144,1536,0.010489777558379702
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,6144,2048,0.024677332904603746
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,2048,0.14312622282240126
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,6144,1536,0.0211173329088423
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,6144,5120,0.04662400152948168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,1536,0.008737777670224508
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,6144,1024,0.017895110779338412
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,1024,0.14476711220211452
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,6144,3584,0.019056888090239633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,6144,768,0.0074720001882976955
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,1024,0.007965332931942409
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,6144,768,0.016520000166363187
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,768,0.14356356196933323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,6144,512,0.006240889016124938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,768,0.007596444752481248
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,512,0.007268444531493717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,6144,512,0.015448888142903646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,512,0.14204800128936768
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,6144,256,0.005211555709441503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,6144,256,0.014642667439248828
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,256,0.14191822210947672
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,6144,128,0.004663111021121343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,7168,0.01982311076588101
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,2048,0.009376889301670922
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,6144,128,0.014135999812020196
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,128,0.13709244463178846
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,6144,64,0.004522666749027041
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,128,0.0069164443347189165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,6144,32,0.004559111264016894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,6144,1024,0.008987555901209513
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,6144,1536,0.147161775165134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,6144,64,0.014088888963063559
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,6144,32,0.013818666338920593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,5120,65536,0.18348088529374865
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,5120,16384,0.11672888861762153
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,5120,65536,0.46160088645087344
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,16384,0.19104799959394667
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,16384,0.03574666712019179
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,5120,12288,0.09081421958075629
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,12288,0.15527289443545872
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,12288,0.0281644430425432
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,5120,10240,0.04046044415897793
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,6144,256,0.0069066666894488865
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,5120,8192,0.03176177872551812
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,10240,0.024460444847742718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,5120,10240,0.07747022310892741
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,65536,0.6703493330213758
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,5120,8192,0.06339733468161689
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,8192,0.020757334099875558
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,8192,0.14494310484992132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,5120,7168,0.02879377868440416
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,5120,6144,0.025133333272404138
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,5120,16384,0.0595946643087599
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,7168,0.019338667392730713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,7168,0.14591556125217015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,5120,5120,0.020875554945733812
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,6144,0.017542221479945712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,6144,0.1415591107474433
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,65536,0.1252008941438463
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,5120,4096,0.01647733317481147
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,5120,12288,0.038147555457221136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,5120,0.015791111522250705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,5120,0.14525689019097224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,5120,3584,0.015563555889659457
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,5120,4096,0.035988443427615695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,4096,0.14329155286153158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,4096,0.013697778185208639
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,5120,3072,0.013645333548386892
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,5120,3584,0.03324177861213684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,3584,0.012444444828563266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,5120,3072,0.0306000014146169
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,10240,0.1474613348642985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,5120,2560,0.01202044470442666
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,3072,0.01148799972401725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,2560,0.01036622209681405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,5120,2560,0.027785778045654297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,2560,0.14233511024051243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,5120,2048,0.01089244418674045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,5120,7168,0.05570399761199951
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,2048,0.1417493290371365
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,5120,1536,0.008967999782827165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,5120,6144,0.04912622107399834
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,2048,0.009313777916961247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,5120,1536,0.019886222150590684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,5120,5120,0.042656001117494374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,5120,1024,0.007297777467303806
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,5120,1024,0.016887111796273124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,1024,0.142795549498664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,5120,768,0.006237333433495627
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,1024,0.007627555893527136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,3584,0.14782044622633192
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,5120,768,0.015785778562227886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,5120,512,0.005187555733654234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,3072,0.14443911446465388
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,768,0.14489244090186224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,5120,512,0.014759999182489185
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,512,0.007035555938879649
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,5120,256,0.004243555582231945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,5120,2048,0.023011555274327595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,256,0.14161245028177896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,5120,128,0.004179555508825514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,256,0.006843555718660355
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,5120,128,0.013747555514176687
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,1536,0.008444444172912175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,1536,0.15151466263665095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,5120,64,0.0038764443662431505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,5120,32,0.00416088890698221
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,128,0.007115555306275685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,5120,64,0.013464888764752282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,5120,32,0.013384000294738345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,4096,65536,0.17629333337148032
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,5120,768,0.007330666813585494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,65536,0.08080622222688463
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,4096,65536,0.4197262128194173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,512,0.14376621776156956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,4096,16384,0.10566044516033596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,5120,256,0.014125333891974555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,16384,0.02457422183619605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,4096,12288,0.04035288757748074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,4096,10240,0.03461066550678677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,4096,12288,0.0824435551961263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,12288,0.019552888141738046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,12288,0.14419644408755833
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,65536,0.5328204366895888
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,4096,8192,0.0277128881878323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,5120,128,0.13872799608442518
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,10240,0.01737511157989502
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,10240,0.14464710818396673
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,4096,7168,0.02420177724626329
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,4096,8192,0.057095997863345675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,8192,0.14376978079477945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,8192,0.01516177753607432
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,4096,6144,0.020992888344658747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,4096,7168,0.050628443559010826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,7168,0.14029599560631648
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,4096,5120,0.017445334129863314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,4096,6144,0.044482668240865074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,6144,0.1435342232386271
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,6144,0.013059555656380124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,4096,4096,0.015071999695565967
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,4096,5120,0.03927288783921136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,4096,16384,0.05316444569163852
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,5120,0.14551466041141087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,4096,3584,0.013728000223636627
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,4096,4096,0.03334044416745504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,4096,0.010048000348938836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,4096,0.14621155791812473
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,4096,3072,0.01219466659757826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,16384,0.15648621982998318
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,3584,0.1451173358493381
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,4096,3072,0.0284222231970893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,4096,2560,0.010947555303573608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,3072,0.009441778063774109
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,3072,0.14881600273980036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,4096,10240,0.07051733467313978
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,4096,2560,0.02590222160021464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,4096,2048,0.009701333112186855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,2560,0.008372444245550368
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,4096,2048,0.021274665991465252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,2048,0.00797333319981893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,4096,1536,0.00794400026400884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,7168,0.014249778456158109
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,4096,1536,0.018596443865034316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,1536,0.00739822205570009
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,4096,1024,0.006272000157170826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,1536,0.14178400569491917
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,5120,0.011745777395036487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,4096,768,0.005247999810510212
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,4096,1024,0.016184889607959323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,1024,0.14975644482506648
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,4096,768,0.015156444576051501
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,4096,3584,0.030826667944590252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,768,0.006664000037643645
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,4096,512,0.004518222063779831
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,3584,0.009415999882751042
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,4096,256,0.004219555606444676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,4096,512,0.014430221584108142
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,512,0.006531555619504716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,512,0.15029866165584987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,4096,256,0.01348088930050532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,4096,128,0.0038337777886125776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,256,0.006127111199829314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,256,0.14509956041971842
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,2560,0.1443084478378296
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,4096,64,0.003556444413132138
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,4096,128,0.01311822235584259
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,128,0.005941333456171884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,128,0.1406773328781128
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,2048,0.14431022273169622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,4096,64,0.013131555583741931
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,4096,32,0.013398222625255585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3584,65536,0.14921511544121638
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3584,16384,0.048327111535602145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,4096,1024,0.0069857777820693115
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,65536,0.07833955685297649
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3584,65536,0.41492178705003524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3584,16384,0.10308000114228989
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,16384,0.023582221733199224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,4096,768,0.14066755771636963
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,16384,0.15600444210900202
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3584,10240,0.03002133303218418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,12288,0.019310222731696237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,12288,0.14404355155097112
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,65536,0.5302800072564019
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3584,10240,0.06931289037068684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3584,8192,0.024803555674023096
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,10240,0.14113777213626438
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,10240,0.017107556263605755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3584,7168,0.020048000746303134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,8192,0.014997333288192749
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3584,8192,0.05583733320236206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,8192,0.14503378338283962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3584,6144,0.01872977779971229
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3584,7168,0.049287110567092896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,7168,0.013849777479966482
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,7168,0.14032799667782253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3584,6144,0.04416266745991177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,6144,0.01278133359220293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,6144,0.145087997118632
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3584,5120,0.038564443588256836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3584,4096,0.013658666776286231
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,5120,0.14078844918145075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,5120,0.011366222467687396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3584,12288,0.032638221979141235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3584,3584,0.01239466667175293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,4096,0.01000088867213991
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3584,4096,0.032778667079077825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3584,12288,0.08078755272759332
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3584,3072,0.011437333292431302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3584,3584,0.030188444587919448
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,3584,0.009389333426952362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,3584,0.146506667137146
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3584,2560,0.010215111076831818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3584,3072,0.02764355474048191
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,3072,0.008981333010726506
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,3072,0.14362222618526882
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3584,2560,0.02534666657447815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,4096,32,0.0038497779104444715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,2560,0.14973689450158015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,2560,0.008453333543406593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3584,1536,0.007684444387753804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3584,2048,0.021253334151373968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,2048,0.007732444339328342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3584,1536,0.018228444788191054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3584,1024,0.006017777654859755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3584,5120,0.016158221496476066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3584,1024,0.016016888949606154
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,1024,0.0069155556460221606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,1024,0.14575733078850642
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3584,768,0.015048000547620984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,768,0.14803644021352133
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3584,512,0.004864000197913912
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,4096,0.14091289043426514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,768,0.006575111299753189
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3584,512,0.01407911048995124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3584,256,0.0042275554604000514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,512,0.006248000181383557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,512,0.14953245057000056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,256,0.005993777679072485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3584,128,0.0038328886859946777
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3584,128,0.013079110946920184
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,2048,0.14613866806030273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,1536,0.14810933007134333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3584,64,0.003858666867017746
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,128,0.0059368887709246735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3584,32,0.003866666720973121
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3584,768,0.005547555370463266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3584,64,0.01310488912794325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3584,32,0.013401777380042605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3072,65536,0.1417893303765191
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,65536,0.07754577530754937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3072,65536,0.37611111005147296
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3072,16384,0.03910400138960944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3584,256,0.013429332938459186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,256,0.14007733927832708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3072,16384,0.0949911144044664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3072,12288,0.032953778902689614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,16384,0.023199111223220825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,65536,0.5268230968051487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,12288,0.14011111524369982
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,3584,1536,0.007314666277832455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3072,10240,0.02748888896571265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,3584,128,0.13489422533247206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,12288,0.018982221682866413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3072,8192,0.022481777601771887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3072,10240,0.06288444333606295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,10240,0.14031378428141275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3072,8192,0.05098933312628004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,8192,0.014753777119848462
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3072,7168,0.01976799964904785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,8192,0.1456800036960178
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,7168,0.013423999978436364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3072,7168,0.04555911156866285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,7168,0.14514844947391087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3072,6144,0.0174346665541331
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,6144,0.012496000362767113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,6144,0.14465155866410997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3584,2048,0.009002666506502364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3072,5120,0.014927999840842353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,16384,0.1562159988615248
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3072,5120,0.03586577706866794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3072,4096,0.012703999876976013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,5120,0.1462551090452406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3072,12288,0.07406310902701484
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3072,4096,0.030234667989942763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3072,3584,0.011443555355072021
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,4096,0.143669327100118
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3072,3584,0.028203556934992473
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,10240,0.017080000705189176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,3584,0.009521777431170145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,3584,0.15189599990844727
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3072,3072,0.010638222098350525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3072,2560,0.009352888498041365
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3072,3072,0.02549955579969618
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,3072,0.14352888531155056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3072,2560,0.02251911163330078
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3072,2048,0.007993777592976889
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,2560,0.008385777473449707
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,2560,0.14082133769989014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3072,2048,0.019337778290112812
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3072,1536,0.006806222101052602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,2048,0.007873777714040544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,2048,0.14432266023423937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3072,6144,0.04032977753215366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3072,1024,0.005934222290913264
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3072,1536,0.017430222696728177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,1536,0.007368889119890001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,1536,0.14097511768341064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3072,768,0.005223999834722943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,5120,0.011198221809334226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3072,1024,0.015170666906568738
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,1024,0.0069155556460221606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,1024,0.14627733495500353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3072,512,0.004514666481150521
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3072,768,0.014136888914638095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,4096,0.009768000079525841
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,768,0.14517866240607366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3072,256,0.004230222354332606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3072,512,0.013801777528391944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,512,0.006274666637182236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,512,0.14312711026933458
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3072,128,0.0038364442686239877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3072,256,0.013074666261672974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3072,64,0.003855111284388436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,256,0.005921777751710679
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3072,128,0.013060444758998023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,3072,32,0.003846222327815162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,128,0.13274044460720485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,3072,0.008814222282833522
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3072,64,0.012503999802801343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,3072,32,0.012763555679056378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2560,65536,0.1276328828599718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2560,16384,0.03788177834616767
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,65536,0.07614399989446004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2560,16384,0.09349777963426377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2560,65536,0.37216001086764866
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,16384,0.1547982162899441
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2560,12288,0.028704888290829126
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,16384,0.022664000590642292
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2560,10240,0.024772443705134924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,12288,0.01869777838389079
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,12288,0.14281333817376032
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2560,12288,0.07289955351087782
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,65536,0.5264870855543349
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,768,0.006627555522653792
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,10240,0.01681511104106903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,10240,0.146525330013699
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2560,7168,0.017299556069903903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2560,8192,0.04902044600910611
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,8192,0.014276444911956787
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,8192,0.14158756203121609
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,3072,256,0.14556445015801325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2560,6144,0.015464888678656684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2560,7168,0.04433244466781616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,7168,0.0129439996348487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,7168,0.1510897742377387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,6144,0.14099733034769693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,6144,0.012094222009181976
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2560,5120,0.03537688983811273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,5120,0.1413066652086046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2560,4096,0.011993777420785693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,5120,0.010877333581447601
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2560,4096,0.03046755658255683
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,4096,0.1501493321524726
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2560,3584,0.011040888726711273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2560,10240,0.06146933635075887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2560,3584,0.027834667099846735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,3584,0.14395377371046278
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2560,3072,0.00999022192425198
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,3072,128,0.005958222266700532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2560,5120,0.01382844481203291
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2560,3072,0.02561155623859829
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2560,6144,0.0406275557147132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,3072,0.14785688453250462
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2560,2560,0.02235200007756551
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,2560,0.14734221829308405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,2560,0.00832266691658232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2560,2048,0.01926488843229082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,4096,0.009645333720578088
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2560,8192,0.019592000378502738
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2560,1536,0.006676444576846228
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,2048,0.1452044381035699
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,2048,0.007655111451943715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2560,1536,0.01720711092154185
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,1536,0.007251555720965068
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,3584,0.009375111096435124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2560,1024,0.005590222362014983
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2560,1024,0.01513066722287072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,1024,0.15074933899773493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2560,768,0.004579555657174853
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,3072,0.00867733359336853
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2560,768,0.014142221874660917
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,768,0.14098755518595377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2560,512,0.004274666723277834
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,768,0.006567111031876669
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2560,256,0.0038417776425679526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2560,512,0.013763555222087435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2560,2048,0.0075911109646161394
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,512,0.13886666297912598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2560,128,0.00349866681628757
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2560,256,0.013091555900043912
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,256,0.14514488644070095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2560,128,0.012724444270133972
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2560,64,0.0032088889016045462
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,128,0.005930666708283954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,128,0.13489688767327204
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2560,32,0.0034977777136696708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2560,64,0.012437333663304647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2560,32,0.012748444245921241
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2048,65536,0.098380446434021
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,2560,1536,0.1476915544933743
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2048,16384,0.03497777713669671
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,1024,0.006710222197903528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,65536,0.07462666432062785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2560,2560,0.00923466682434082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2048,12288,0.02711822258101569
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2048,65536,0.3351617654164632
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,16384,0.02232711182700263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,16384,0.15264799859788683
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2048,10240,0.022701332966486614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2048,12288,0.06508444415198432
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,12288,0.01835466590192583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,512,0.00627377794848548
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,65536,0.5096657541063097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2048,8192,0.019055111540688407
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,10240,0.016142222616407607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,10240,0.1425386667251587
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,2560,256,0.005904888941182031
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2048,8192,0.044453332821528115
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,8192,0.1495315498775906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,8192,0.013444444371594323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2048,6144,0.01512977812025282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2048,7168,0.04115644428465102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,7168,0.012810666528013019
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,7168,0.1425973309410943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2048,5120,0.013428444663683573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2048,6144,0.03603377938270569
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,6144,0.14572532971700033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,6144,0.011723555624485016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2048,4096,0.011707555916574268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2048,5120,0.03219377663400438
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,5120,0.14523555172814265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,5120,0.01072622256146537
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2048,3584,0.010659555594126383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2048,4096,0.027285334136750963
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,4096,0.009639110830095079
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2048,16384,0.08023111025492351
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2048,3072,0.009319110876984065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2048,3584,0.025244444608688354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,3584,0.15025333563486734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,3584,0.009020444419648912
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2048,2560,0.008329778081840938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2048,3072,0.022740445203251306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,3072,0.00869955536392
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,3072,0.14701688289642334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2048,10240,0.054486221737331815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2048,2048,0.006933333145247565
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2048,2560,0.020627554919984605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,2560,0.00833777752187517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,2560,0.14699199464586046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2048,1536,0.006067555397748947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,2048,0.14740444554222956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,2048,0.007631999750932057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2048,1536,0.01623288955953386
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2048,1024,0.005025777965784073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,1536,0.007299555672539606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,1536,0.14659999476538763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2048,1024,0.014477334088749357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,1024,0.006625777731339137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2048,768,0.004564444637960858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,1024,0.15132889482710096
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2048,768,0.013750221994188098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,768,0.14563111464182535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2048,512,0.004176889028814104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,4096,0.14200977484385172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2048,512,0.013173333472675748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,512,0.006224000205596288
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,512,0.15051644378238255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,12288,0.1414746708340115
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2048,256,0.003867555409669876
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2048,128,0.0038613333470291565
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2048,256,0.012799111505349478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,256,0.1454764472113715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2048,128,0.012699555191728802
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2048,7168,0.017039111918873258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2048,64,0.003532444437344869
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2048,2048,0.01813155578242408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,128,0.005943999936183293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,2048,128,0.13528888755374485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2048,64,0.012126222252845764
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1536,65536,0.08205688661999173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,2048,32,0.012392000191741519
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1536,16384,0.02784444557295905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,65536,0.07333244217766656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1536,65536,0.3253653314378527
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1536,16384,0.08242311080296834
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1536,12288,0.019627556204795837
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,16384,0.022223111655977037
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,16384,0.1516817808151245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,768,0.006595555692911148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1536,10240,0.017031111650996737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1536,12288,0.06350666946834989
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,12288,0.01773066653145684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,12288,0.14487022823757595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,65536,0.5065129068162706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1536,10240,0.05177866750293308
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,10240,0.14325066407521567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,10240,0.015494222442309061
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1536,7168,0.013428444663683573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1536,8192,0.04435999857054817
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,8192,0.14301689465840658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,8192,0.013437333206335703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1536,6144,0.012327111429638334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,2048,256,0.005943111247486538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1536,7168,0.039792001247406006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,7168,0.14528089099460179
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,7168,0.01236355553070704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1536,5120,0.010771555205186209
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1536,6144,0.03536533315976461
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,6144,0.011523555550310345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,2048,32,0.003847111016511917
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1536,4096,0.009198221895429824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,6144,0.14215911759270564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1536,5120,0.031222220924165513
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1536,3584,0.008632889224423302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,5120,0.010682666467295753
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,5120,0.14419378174675837
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1536,3072,0.007592889169851939
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,4096,0.009659555223253038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,4096,0.14515822463565403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,3584,0.008960000342792934
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,3584,0.14530311690436468
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1536,2560,0.006958222223652734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1536,3072,0.02254933284388648
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,3072,0.00867555538813273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,3072,0.1520222160551283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1536,2048,0.006212444355090459
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1536,2560,0.020475554797384474
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,2560,0.008260444634490544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,2560,0.140956441561381
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1536,1536,0.005520888914664586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,2048,0.14613866806030273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1536,1536,0.016156444946924847
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1536,1024,0.004565333326657613
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,1536,0.1447075605392456
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1536,8192,0.014255111416180929
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,1536,0.007300444775157505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1536,768,0.0042071110672420925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,1024,0.006611555400821898
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,1024,0.14736266930898032
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1536,768,0.013785777820481194
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1536,512,0.003916444463862313
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,768,0.1441377798716227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1536,512,0.013111111190583972
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,512,0.14250932799445257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1536,4096,0.02642311155796051
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,512,0.006264888991912206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1536,3584,0.02399377855989668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1536,256,0.012720000412729053
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,256,0.14574310514662
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1536,128,0.0031848889258172777
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,256,0.005933333188295364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1536,2048,0.017584888471497428
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,2048,0.007623111208279927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1536,64,0.003220444545149803
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,128,0.005962666538026597
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,1536,128,0.13277955849965414
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1536,32,0.0031973332580592898
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1536,64,0.012053333222866058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1024,65536,0.05317955546908908
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1536,32,0.012074666718641916
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1536,1024,0.014472888575659858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1024,16384,0.01936177743805779
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,65536,0.07162755727767944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1024,16384,0.07214311096403334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1024,65536,0.28740088144938153
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1024,12288,0.018820444742838543
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,16384,0.14974578221638998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,16384,0.021193777521451313
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1536,256,0.0034640000926123727
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,65536,0.5058506859673394
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1024,10240,0.01633066601223416
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1024,12288,0.0543084442615509
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,12288,0.14140355587005615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,12288,0.016487111647923786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1024,8192,0.014117333624098035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,10240,0.014999111493428549
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1024,10240,0.04592533244027031
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1024,7168,0.01278222186697854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1024,8192,0.038182222180896334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,8192,0.14060178067949083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,8192,0.013089777694808112
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1024,6144,0.011459555890825061
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1024,7168,0.034359110726250544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,7168,0.14325688944922552
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,7168,0.012152888708644442
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1024,5120,0.010024888647927178
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1024,6144,0.03120888935195075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,6144,0.011414222419261932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,6144,0.1495937771267361
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1024,4096,0.008774222599135505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,5120,0.14643643962012395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,5120,0.010671999719407825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1024,3584,0.007616889145639207
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1024,4096,0.023451555106374953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,4096,0.1405857801437378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,4096,0.00941066692272822
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1024,3072,0.006966222491529252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1024,3584,0.02201066745652093
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,3584,0.009008888569143083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,3584,0.14983556005689833
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1024,2560,0.006598222172922558
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1024,3072,0.020279111133681405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,3072,0.14374666743808323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1024,2560,0.018558222386572096
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,2560,0.14451644155714247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,2560,0.007997333175606197
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,10240,0.14346222082773843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1024,2048,0.016512889001104567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,2048,0.14356356196933323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1024,1536,0.005021333280536863
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1024,1536,0.01512977812025282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,1536,0.14608089129130045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1024,1024,0.004455111092991299
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,1536,0.007267555428875818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1024,5120,0.02823644545343187
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1024,1024,0.013750221994188098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,1024,0.006628444625271692
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,1536,768,0.0063031112982167145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1024,768,0.004190222256713443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,1024,0.1439244482252333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1024,512,0.0035706667436493766
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1024,768,0.01310755560795466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,768,0.14716711309221056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1024,512,0.01256533298227522
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1024,256,0.0031742221779293488
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,3072,0.008403555386596257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,512,0.006247111078765657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,512,0.14054755369822183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1024,128,0.003162666741344664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1024,256,0.0123857781291008
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1024,2048,0.005803555664088991
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,256,0.005950222412745158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,256,0.1424328883488973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1024,64,0.0028195555011431375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1024,128,0.01202933324707879
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,1024,32,0.0031724443866146933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,128,0.005928888916969299
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,1024,128,0.13619021574656168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,2048,0.007649777664078607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1024,64,0.011751111182901593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,768,65536,0.04495911134613884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,768,16384,0.017433777451515198
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,65536,0.07033777899212308
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,768,16384,0.07067022058698866
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,16384,0.14976355764600965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,768,65536,0.28548532062106663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,768,12288,0.01585422290696038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,16384,0.020837333467271592
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,768,12288,0.055027554432551064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,12288,0.14990222454071045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,65536,0.4992408752441406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,12288,0.016584000653690763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,768,10240,0.014715555641386243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,768,8192,0.013239999612172445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,1024,768,0.006256888724035687
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,10240,0.014890667464998035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,10240,0.14414932992723253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,768,7168,0.012545777691735161
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,768,8192,0.03763022356563144
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,8192,0.013051555388503604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,8192,0.14633954895867243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,768,6144,0.011167999770906238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,768,7168,0.033943110042148165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,7168,0.1414382192823622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,7168,0.012130666938092975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,768,5120,0.009523555636405945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,768,6144,0.030335999197430078
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,6144,0.011400889191362592
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,6144,0.1439324484931098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,768,4096,0.008007110820876228
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1024,32,0.011853333148691388
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,768,5120,0.027097778187857732
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,5120,0.14519200060102674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,5120,0.01038933296998342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,768,3584,0.007489778101444244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,768,4096,0.023623999622133043
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,4096,0.009362666971153682
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,4096,0.13993777169121638
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,768,3584,0.021760000122918025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,3584,0.14294755458831787
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,768,3072,0.020261334048377145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,3072,0.14083288775549993
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,768,2560,0.006259555617968242
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,3072,0.008608000146018134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,2560,0.008037333687146505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,768,10240,0.045257776975631714
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,2560,0.14221510622236463
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,768,2048,0.005600888695981767
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,768,2048,0.016176889340082806
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,2048,0.007627555893527136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,2048,0.1420960028966268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,768,1536,0.005261333452330695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,1536,0.007095110913117726
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,768,1536,0.015101333459218344
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,1536,0.1414533323711819
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,768,1024,0.004567111117972268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,1536,128,0.012425777812798819
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,768,768,0.004214222232500712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,1024,0.006772444479995304
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,768,1024,0.013440888788965015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,1024,0.1407973368962606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,768,512,0.0038604442444112566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,768,768,0.01309511148267322
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,768,0.00628266649113761
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,768,0.15124000443352595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,768,256,0.0035173334181308746
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,3584,0.008972444468074376
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,512,0.00628266649113761
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,768,128,0.0031351111829280853
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,256,0.005972444597217772
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,256,0.15157332685258654
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,768,128,0.012044444680213928
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,768,2560,0.01890222231547038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,768,64,0.002855111120475663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,768,128,0.005943999936183293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,128,0.13783733050028482
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,768,32,0.003145777723855443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,512,65536,0.035183111826578774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,768,64,0.011756444142924415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,768,32,0.011744000017642975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,512,16384,0.014441778262456259
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,65536,0.06844800048404269
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,512,65536,0.2845377657148573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,512,16384,0.07005866368611653
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,512,12288,0.012378666963842181
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,16384,0.019902222686343722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,65536,0.5009519788953993
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,512,12288,0.05274933245446947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,12288,0.1433466672897339
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,512,10240,0.014449778530332776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,12288,0.016194666425387066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,768,512,0.012410666379663678
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,512,8192,0.013419555293189155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,768,512,0.14506399631500244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,10240,0.014818666709793938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,512,10240,0.045462220907211304
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,10240,0.1499511135949029
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,768,256,0.012058667010731168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,512,7168,0.012406222522258759
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,512,8192,0.03803822067048814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,8192,0.01277955538696713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,512,6144,0.01072533345884747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,512,7168,0.03358577688535055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,7168,0.012033777932325998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,7168,0.1482328838772244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,512,5120,0.009293333523803288
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,512,6144,0.030125333203209773
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,6144,0.14116533597310385
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,512,5120,0.02700533303949568
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,5120,0.14471199777391222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,512,4096,0.008003555238246918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,5120,0.010488889283604093
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,512,4096,0.023349333140585158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,4096,0.14817689524756536
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,512,3584,0.007289778027269576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,16384,0.14877422650655112
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,512,3584,0.021617778473430212
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,512,3072,0.0069066666894488865
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,3584,0.14679733912150064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,3584,0.008984888593355814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,768,3072,0.006978666616810693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,512,2560,0.005968889014588461
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,512,3072,0.019894222418467205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,3072,0.008340444829728868
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,3072,0.14332000414530435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,512,2048,0.00554044461912579
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,512,2560,0.018582221534517076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,2560,0.00794488853878445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,2560,0.14756622579362658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,8192,0.14480711354149714
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,512,2048,0.016187555260128446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,2048,0.00758755538198683
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,2048,0.1449120044708252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,512,1024,0.004171555654870139
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,1536,0.1417093276977539
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,1536,0.0070364442136552595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,512,1024,0.013456888496875763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,1024,0.006577777779764599
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,1024,0.1424808899561564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,4096,0.009353777600659264
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,512,768,0.013056889176368713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,768,0.14219021797180176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,768,0.006340444501903322
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,512,512,0.0124195557501581
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,512,0.14127643903096518
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,512,0.006223111102978389
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,512,256,0.012087110843923358
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,512,1536,0.00488711107108328
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,256,0.0058959999846087555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,512,1536,0.015150222513410779
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,512,128,0.0028657778683635923
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,6144,0.011063999599880643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,512,128,0.012118221984969245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,512,64,0.0028444443725877335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,512,768,0.0038293335172865125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,512,32,0.002858666703104973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,256,65536,0.027128888501061335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,512,64,0.012073777616024017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,512,32,0.011886222494973077
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,512,512,0.0035333335399627686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,256,16384,0.011620444556077322
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,65536,0.0677679975827535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,256,65536,0.2836302121480306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,256,16384,0.06911644670698378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,512,256,0.0031964443624019623
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,16384,0.14860978391435412
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,256,12288,0.012873777912722694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,16384,0.0196115556690428
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,65536,0.49656269285413956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,256,10240,0.011361777782440186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,12288,0.016180444094869826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,256,12288,0.051595555411444775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,12288,0.14589599768320718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,256,10240,0.04483555422888862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,10240,0.14425688319736057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,10240,0.014783110883500842
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,256,0.14233333534664577
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,256,7168,0.009648000200589498
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,256,8192,0.03720088799794515
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,8192,0.012784000072214337
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,8192,0.14461066987779406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,256,6144,0.00886933339966668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,256,7168,0.033257779147889875
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,7168,0.012071111136012606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,7168,0.14047733942667642
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,512,128,0.13761511113908556
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,256,5120,0.00832711077398724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,256,6144,0.029824889368481104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,6144,0.1455671124988132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,256,4096,0.007663110891977946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,6144,0.011329777538776398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,256,5120,0.026920888159010146
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,5120,0.010390222072601318
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,5120,0.1461040046479967
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,256,3584,0.0069439998931354946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,4096,0.009347555538018545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,256,4096,0.02316266629430983
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,4096,0.14541067017449275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,256,3072,0.006263999889294307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,256,3584,0.021336000826623704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,3584,0.1478782229953342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,3072,0.008608000146018134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,256,3072,0.019908444748984445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,3072,0.14469333489735922
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,256,2048,0.005209777918126848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,256,2560,0.01853511068556044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,2560,0.007941332956155142
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,2560,0.14118044906192356
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,256,2048,0.01626844373014238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,2048,0.14609955416785345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,256,1536,0.004598222259018156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,2048,0.007626666790909237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,256,8192,0.010388444695207808
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,256,1024,0.0038408889538711975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,1536,0.007237333390447829
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,256,1536,0.01481244464715322
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,1536,0.14788444836934408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,256,768,0.0038720000949170855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,256,1024,0.013412444127930535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,512,128,0.006024888820118374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,1024,0.14140710565778944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,256,512,0.0034871109657817413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,768,0.14067199495103624
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,256,512,0.01239466667175293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,512,0.13961688677469888
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,256,256,0.003145777723855443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,3584,0.008825778133339351
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,512,0.006227555374304454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,256,2560,0.0058746664888328975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,256,128,0.0028124445428450904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,256,256,0.012048000262843238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,256,0.005895111295912001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,256,0.14104178216722277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,256,64,0.0028088889602157804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,256,128,0.011755555868148804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,256,32,0.0028026666906144884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,256,128,0.13262044058905706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,128,65536,0.018053332964579265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,256,64,0.011753777662913004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,256,32,0.011704000333944956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,65536,0.06588177548514472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,128,16384,0.010027555955780877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,1024,0.006588444527652528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,256,768,0.012788444757461548
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,768,0.006252444452709622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,16384,0.01957244508796268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,65536,0.5030595461527506
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,12288,0.016164445214801364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,128,12288,0.0510408878326416
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,12288,0.14688977930280897
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,128,10240,0.009000000026490953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,128,10240,0.043938666582107544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,10240,0.14653689331478542
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,128,8192,0.008349333372380998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,256,128,0.005935110979610019
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,128,8192,0.03660088777542114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,128,7168,0.008014221986134848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,8192,0.012764444781674279
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,128,7168,0.032792889409595065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,7168,0.14368711577521429
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,128,6144,0.007346666521496243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,7168,0.011760888828171624
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,128,12288,0.009001777403884465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,128,6144,0.02947910957866245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,16384,0.14925866656833225
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,6144,0.14639022615220812
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,6144,0.011145778000354767
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,128,5120,0.026594666971100703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,5120,0.1460960043801202
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,128,4096,0.0075706665714581805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,5120,0.01036266651418474
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,10240,0.014543111125628153
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,4096,0.009355555805895064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,128,4096,0.023039110832744177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,4096,0.1460231145222982
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,3584,0.008943110704421997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,128,3584,0.021319111188252766
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,3584,0.14758311377631292
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,128,3072,0.006326222171386083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,128,3072,0.019967110620604623
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,3072,0.008338666624493068
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,3072,0.14858667055765787
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,128,2560,0.005889777921968036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,128,65536,0.28392354647318524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,2560,0.007976888782448238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,2560,0.14766577879587808
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,128,2048,0.005256888767083486
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,128,16384,0.06798488563961454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,128,2048,0.016181333197487723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,128,1536,0.00452888881166776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,128,5120,0.007978666987684038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,2048,0.1453546682993571
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,128,1024,0.0038488888078265717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,1536,0.007228444847795699
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,1536,0.14484443929460314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,128,1024,0.013083555632167392
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,128,768,0.003830222205983268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,1024,0.006654222392373615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,128,3584,0.0069022224181228215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,128,768,0.01274222218328052
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,128,512,0.003208000005947219
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,768,0.14587822225358751
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,8192,0.1513893339369032
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,128,512,0.01205599970287747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,128,256,0.0028764444092909494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,512,0.006248000181383557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,128,256,0.011728888584507836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,128,2560,0.018558222386572096
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,256,0.005893333504597346
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,2048,0.007628444168302748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,128,128,0.0028515555378463534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,128,128,0.011672000090281168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,128,1536,0.014840000205569796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,128,64,0.0025128889828920364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,128,0.005928888916969299
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,128,32,0.0025271111064487034
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,128,64,0.011770666473441653
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,64,65536,0.016606221596399944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,128,32,0.011405333048767514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,512,128,768,0.006291555447710886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,64,16384,0.009671111073758867
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,64,12288,0.008638222184446123
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,512,0.14426755905151367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,64,16384,0.06877511077457003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,64,65536,0.28356266021728516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,64,10240,0.009024000002278222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,64,12288,0.05132088727421231
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,64,8192,0.008201777935028076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,64,7168,0.00793599999613232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,64,6144,0.007307555940416124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,256,0.14876977602640787
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,64,10240,0.042288889487584434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,64,8192,0.03598489032851325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,64,5120,0.007977777885066139
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,64,7168,0.03265422251489427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,64,4096,0.007285333342022366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,64,3584,0.0069777775141927935
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,64,3072,0.006654222392373615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,64,6144,0.029503110382292006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,64,2560,0.006268444574541516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,64,5120,0.02648888859483931
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,64,4096,0.02275466587808397
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,64,2048,0.005893333504597346
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,64,3584,0.02126400007141961
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,64,3072,0.019729778170585632
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,64,1024,0.0038684445122877755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,64,2048,0.015824000040690105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,64,1536,0.014564444621404013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,64,768,0.003536000019974179
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,64,512,0.003244444520937072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,64,1024,0.013443555268976422
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,1024,0.14244088861677381
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,64,256,0.0028657778683635923
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,64,768,0.012409778104888068
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,64,128,0.002830222249031067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,64,512,0.012070222033394707
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,64,64,0.0028488888508743713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,64,32,0.0026168889469570587
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,64,128,0.011389333340856763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,64,64,0.011389333340856763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,32,65536,0.028777778148651123
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,64,32,0.011720000041855706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,32,16384,0.06831466489368014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,32,65536,0.2832186751895481
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,32,10240,0.007986666427718269
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,32,12288,0.0503768887784746
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,32,8192,0.0083137775460879
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,32,10240,0.043466667334238686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,64,1536,0.004512888689835866
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,32,7168,0.007982222570313348
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,64,2560,0.018237334158685472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,32,6144,0.007627555893527136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,32,5120,0.006959999601046245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,32,8192,0.035601778162850276
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,32,4096,0.007631111476156447
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,32,7168,0.03279733326699998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,32,6144,0.029514667060640123
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,32,5120,0.026378666361172993
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,32,3072,0.006945778098371293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,32,4096,0.022720888257026672
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,32,2560,0.006966222491529252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,32,3584,0.02126755482620663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,64,256,0.012056888805495368
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,32,2048,0.006246222390068903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,32,1536,0.004528000122971005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,32,3072,0.01962844365172916
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,32,1024,0.0038604442444112566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,32,2560,0.01828799976242913
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,32,2048,0.016152888536453247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,32,768,0.003532444437344869
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,32,1536,0.014767999450365702
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,32,512,0.003165333428316646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,32,256,0.0028453332682450614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,32,1024,0.0129120002190272
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,32,768,0.013059555656380124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,32,12288,0.008461332983440822
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,32,512,0.012079999678664737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,32,64,0.0028088889602157804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,32,256,0.012073777616024017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,32,32,0.002612444468670421
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,32,128,0.011759111450778114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,32,64,0.011400889191362592
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,512,32,32,0.011408888631396823
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,65536,16384,0.4920141961839464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,32,3584,0.007245333658324347
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,65536,16384,0.5389128790961372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,65536,16384,0.2275315655602349
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,32,16384,0.010050666828950247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,65536,12288,0.3992462158203125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,512,32,128,0.0028400001012616684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,65536,10240,0.29360000292460126
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,65536,12288,0.17683200041453043
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,65536,10240,0.3383164405822754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,65536,8192,0.24258221520317924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,65536,10240,0.1494942241244846
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,65536,12288,0.38408178753323025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,65536,8192,0.27737511528862846
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,65536,16384,0.6312284469604492
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,65536,12288,0.4852577845255534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,65536,7168,0.2134257819917467
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,65536,7168,0.25854844517178005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,65536,10240,0.40508177545335555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,65536,6144,0.18002577622731528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,65536,8192,0.34170932239956325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,65536,6144,0.2313297854529487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,65536,5120,0.19945422808329263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,65536,7168,0.3115982214609782
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,65536,6144,0.09209955400890774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,65536,4096,0.12423466311560737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,65536,5120,0.17902578247918022
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,65536,5120,0.08014400137795342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,65536,6144,0.2647751172383626
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,65536,4096,0.14575821823543972
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,65536,3584,0.12190311484866673
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,65536,8192,0.12155644098917644
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,65536,5120,0.2357849015129937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,65536,4096,0.06640266709857516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,65536,7168,0.10557422373029922
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,65536,3072,0.12113600307040745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,65536,3584,0.057730668120914035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,65536,4096,0.19504977597130668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,65536,3072,0.10958488782246907
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,65536,2560,0.08464799986945258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,65536,3072,0.051254222790400185
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,65536,3584,0.18692266941070557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,65536,2560,0.09581777784559463
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,65536,2048,0.07984089189105563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,65536,2560,0.04536799920929802
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,65536,3072,0.1610328886244032
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,65536,2560,0.14583555857340494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,65536,1536,0.05554222067197164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,65536,2048,0.0823884473906623
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,65536,2048,0.03864088985655043
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,65536,1024,0.03700444433424208
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,65536,1536,0.06553333335452609
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,65536,2048,0.14186044534047446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,65536,1536,0.030638221237394545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,65536,768,0.0343795551194085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,65536,1536,0.14406932724846735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,65536,1024,0.05266133281919691
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,65536,512,0.023544000254737005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,65536,1024,0.1475084490246243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,65536,768,0.04435555471314324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,65536,768,0.021906667285495337
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,65536,1024,0.024471999870406255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,65536,256,0.01661688917213016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,65536,512,0.038784000608656145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,65536,512,0.14623733361562094
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,65536,512,0.019723556107944913
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,65536,128,0.0177831103404363
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,65536,256,0.01831377711560991
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,65536,256,0.14353511068556044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,65536,128,0.03191555539766947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,65536,64,0.014916444818178812
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,65536,128,0.13867910703023276
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,65536,128,0.018194667167133756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,65536,64,0.03159911102718777
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,65536,32,0.03159111075931125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,16384,65536,0.4880702230665419
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,65536,768,0.1415146721733941
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,16384,65536,0.688486205206977
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,16384,16384,0.1228391064537896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,65536,256,0.03356711069742838
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,16384,16384,0.17513067192501494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,16384,65536,0.2519377867380778
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,16384,12288,0.09573155641555786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,16384,12288,0.13509421878390843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,16384,16384,0.24649156464470756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,16384,10240,0.08605600065655178
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,16384,12288,0.05123555660247803
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,16384,12288,0.19506222671932646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,16384,10240,0.11121422714657253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,16384,8192,0.06608266962899102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,16384,10240,0.04388977752791511
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,65536,3584,0.12791200478871664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,16384,10240,0.17496355374654135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,16384,7168,0.06355644596947564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,65536,32,0.015098666151364645
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,16384,8192,0.09140266974767049
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,16384,8192,0.03616177704599168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,16384,8192,0.1504462162653605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,16384,65536,0.8898631201850044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,16384,7168,0.0827946662902832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,16384,6144,0.05774488714006212
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,16384,5120,0.04664533336957296
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,16384,6144,0.07255555523766412
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,16384,7168,0.1441297796037462
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,16384,6144,0.02903911140229967
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,16384,6144,0.1470524470011393
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,16384,5120,0.062124444378746875
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,16384,4096,0.03954844342337714
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,16384,5120,0.14782933394114176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,16384,4096,0.052758223480648465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,16384,3584,0.035283555587132774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,16384,4096,0.02143200072977278
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,16384,4096,0.1473040050930447
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,16384,3584,0.04839555422465006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,16384,3072,0.03113244308365716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,16384,3584,0.14800800217522517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,16384,3584,0.01980622278319465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,16384,3072,0.0421635541650984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,16384,3072,0.017820444371965196
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,16384,3072,0.14711111121707493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,16384,16384,0.06601955493291219
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,16384,2560,0.03721155391799079
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,16384,7168,0.03270666797955831
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,16384,2560,0.1466933356391059
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,16384,2048,0.021555556191338435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,16384,2048,0.01462311049302419
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,16384,2048,0.14768266677856445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,16384,1536,0.01700177788734436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,16384,5120,0.0255022214518653
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,16384,1536,0.026826666461096868
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,16384,1536,0.14759111404418945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,16384,1024,0.013099555340078143
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,16384,1024,0.022962666220135156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,16384,1024,0.14613421758015951
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,16384,768,0.011911110745535957
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,16384,1024,0.010737777584128909
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,16384,768,0.020215999748971727
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,16384,2560,0.024422221713595923
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,16384,768,0.14787556065453425
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,16384,512,0.009638222555319468
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,16384,768,0.009747555686367882
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,16384,2560,0.016618667377365958
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,16384,512,0.018519111805491976
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,16384,256,0.007616889145639207
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,16384,512,0.009026666482289633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,16384,512,0.14097155465020075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,16384,128,0.006670222100284364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,16384,2048,0.031552000178231135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,16384,256,0.016901334126790363
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,16384,256,0.1508773300382826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,16384,64,0.00628977765639623
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,16384,128,0.016528889536857605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,16384,32,0.006568000134494569
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,16384,128,0.00870666652917862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,16384,128,0.13391555680169
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,16384,1536,0.01295377810796102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,16384,64,0.016161777906947665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,16384,32,0.016941333810488384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,12288,16384,0.10551022158728705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,12288,65536,0.36435556411743164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,12288,16384,0.14398311244116888
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,12288,65536,0.22118043899536133
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,12288,65536,0.5641688770718044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,12288,12288,0.07946222358279757
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,12288,16384,0.21507110860612658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,12288,12288,0.10923377672831218
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,12288,12288,0.043997334109412305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,16384,256,0.008619555168681675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,12288,12288,0.16851645045810273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,12288,10240,0.07589333587222628
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,12288,10240,0.0924160016907586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,12288,8192,0.05386222071117825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,12288,10240,0.03676977753639221
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,12288,10240,0.15148622459835476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,12288,8192,0.07687199778027005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,12288,8192,0.028984887732399836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,12288,8192,0.15285421742333308
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,12288,7168,0.06868088907665677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,12288,6144,0.04362933172119988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,12288,7168,0.14727911684248182
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,12288,7168,0.025926222403844196
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,12288,5120,0.03924622138341268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,12288,6144,0.023880889018376667
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,12288,6144,0.15189866224924722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,12288,65536,0.7448622385660807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,12288,5120,0.05168888966242472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,12288,4096,0.03245688809288873
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,12288,5120,0.14628267288208008
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,12288,4096,0.04302933481004503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,12288,3584,0.02572622232966953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,12288,4096,0.018242667118708294
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,12288,4096,0.14364088906182185
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,12288,3584,0.03939911060863071
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,12288,3584,0.14703110853830972
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,12288,3584,0.01700088878472646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,12288,3072,0.022737777895397607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,12288,7168,0.0539644459883372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,12288,2560,0.018790221876568265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,12288,3072,0.03454844488037957
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,12288,3072,0.1433280044131809
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,12288,2560,0.030625777112113103
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,12288,2560,0.014192000031471252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,12288,2560,0.1469493309656779
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,12288,2048,0.01682577696111467
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,12288,6144,0.06049244271384346
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,12288,2048,0.027357333236270483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,12288,16384,0.05219111177656385
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,12288,2048,0.14666133456759983
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,12288,1536,0.01367022262679206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,12288,2048,0.013040888640615674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,12288,1024,0.010695999695195092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,12288,5120,0.020938666330443487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,12288,1536,0.010672888822025724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,12288,1536,0.1459475490781996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,12288,768,0.008991111483838823
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,12288,1024,0.020227554771635268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,12288,1024,0.009390222529570261
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,12288,1024,0.151293330722385
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,12288,512,0.008037333687146505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,12288,768,0.0181688881582684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,12288,768,0.14102666907840306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,12288,768,0.00867733359336853
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,12288,256,0.005551110953092575
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,12288,512,0.016565332810084026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,12288,512,0.00832622249921163
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,12288,512,0.145778669251336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,12288,128,0.00489688871635331
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,12288,256,0.015847999188635085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,12288,3072,0.015173332558737861
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,12288,256,0.007931555310885111
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,12288,64,0.00452622233165635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,12288,128,0.015499555402331881
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,12288,128,0.13788089487287733
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,12288,32,0.004598222259018156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,12288,128,0.007983110845088959
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,12288,64,0.01514933341079288
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,12288,32,0.01512977812025282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,10240,65536,0.28682133886549205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,12288,1536,0.023669333921538457
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,10240,16384,0.0958799984720018
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,10240,65536,0.5007146729363335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,10240,65536,0.17829244666629365
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,10240,16384,0.12900266382429335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,10240,12288,0.07474311192830403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,10240,16384,0.04895377821392483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,10240,16384,0.20979200469122994
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,10240,12288,0.09889422522650824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,10240,12288,0.03860266672240363
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,12288,256,0.1523760027355618
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,10240,10240,0.06025244130028618
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,10240,12288,0.1667626698811849
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,10240,8192,0.05004355642530653
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,10240,10240,0.08399199777179295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,10240,10240,0.03317866722742716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,10240,8192,0.07012355327606201
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,10240,10240,0.14941599633958605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,10240,8192,0.14711999893188477
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,10240,8192,0.027961777316199407
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,10240,6144,0.037760890192455716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,10240,7168,0.024953777591387432
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,10240,7168,0.1410977840423584
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,10240,6144,0.05552622344758776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,10240,5120,0.03159822358025445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,10240,65536,0.7276213434007434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,10240,6144,0.02232355541653103
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,10240,6144,0.14635911252763537
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,10240,4096,0.024497777223587036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,10240,5120,0.0466284453868866
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,10240,5120,0.14140710565778944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,10240,5120,0.02016622159216139
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,10240,3584,0.021692444880803425
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,10240,4096,0.14858578311072454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,10240,4096,0.017433777451515198
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,10240,3072,0.0192239996459749
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,10240,3584,0.03677422139379714
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,10240,3584,0.016019556257459853
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,10240,3584,0.14032533433702257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,10240,2560,0.01756444407833947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,10240,3072,0.03194577826393975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,10240,3072,0.014775110615624322
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,10240,3072,0.14126400152842203
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,10240,2048,0.015154666370815702
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,10240,2560,0.013635555903116861
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,10240,2560,0.028691556718614366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,10240,2560,0.14152355988820395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,10240,2048,0.025399110383457605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,10240,1536,0.012522666818565793
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,10240,2048,0.012072000238630505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,10240,2048,0.14762843979729548
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,10240,7168,0.06345511145061918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,10240,1024,0.010429333481523724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,10240,1536,0.022231111923853557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,10240,1536,0.14195911089579263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,10240,1536,0.010325333310498131
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,10240,768,0.00869422240389718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,10240,1024,0.01905866629547543
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,10240,1024,0.009010666774378883
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,10240,1024,0.1474319961335924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,10240,512,0.007249777515729268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,10240,768,0.017162667380438913
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,10240,768,0.14381778240203857
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,10240,256,0.005884444548024072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,10240,4096,0.04012711180580987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,10240,512,0.015853333804342482
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,10240,512,0.00798044436507755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,10240,512,0.1436097754372491
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,10240,128,0.0052062223354975385
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,10240,256,0.015159999330838522
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,10240,256,0.007656888829337225
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,10240,256,0.1458968851301405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,512,128,128,0.14106578297085232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,10240,128,0.014442667365074158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,10240,32,0.005611555443869696
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,10240,128,0.007625777688291337
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,10240,128,0.13625066810184056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,10240,64,0.014496889379289416
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,10240,32,0.01516266663869222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,10240,7168,0.04468799961937798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,8192,65536,0.2428684499528673
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,8192,16384,0.06586755646599664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,8192,65536,0.4485137727525499
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,8192,65536,0.13417332702212864
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,8192,16384,0.11639377805921768
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,8192,12288,0.05036355720625984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,8192,16384,0.17694400416480172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,8192,16384,0.03861066699028015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,10240,768,0.008412444757090675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,8192,12288,0.08965510792202419
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,8192,10240,0.04205244448449877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,8192,12288,0.02994133366478814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,8192,12288,0.14191199673546687
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,8192,8192,0.0414266652531094
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,8192,10240,0.02623111175166236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,8192,10240,0.14213511678907606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,10240,64,0.004855111241340637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,8192,7168,0.03223911258909438
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,8192,8192,0.06394133302900526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,8192,8192,0.1518497731950548
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,8192,8192,0.022278222772810195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,8192,6144,0.030757334497239854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,8192,7168,0.05683910846710205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,8192,7168,0.020216888851589628
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,8192,65536,0.589171568552653
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,8192,5120,0.025058666865030926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,8192,6144,0.050548444191614784
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,8192,6144,0.018041777941915724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,8192,6144,0.149853335486518
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,8192,5120,0.04308177696333992
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,8192,5120,0.016407110624843173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,8192,5120,0.14523643917507595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,8192,3584,0.01862044466866387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,8192,4096,0.03687289026048448
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,8192,4096,0.014544889330863953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,8192,4096,0.14875022570292154
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,8192,3072,0.016692444682121277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,8192,3584,0.033703111939960055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,8192,3584,0.14893599351247153
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,8192,3584,0.013464888764752282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,8192,2560,0.014535999960369535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,8192,3072,0.02948088778389825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,8192,3072,0.1417671177122328
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,8192,2048,0.012759110993809171
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,8192,2560,0.026735999517970618
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,8192,2560,0.011397333608733283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,8192,10240,0.07729778024885389
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,8192,2048,0.023696889479955036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,8192,1536,0.011007111105653973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,8192,2048,0.14768621656629774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,8192,2048,0.010096888575288985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,8192,1024,0.008954666554927826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,8192,1536,0.020785777105225455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,8192,1536,0.008997333546479544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,8192,1536,0.14500444465213352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,8192,768,0.007323555648326874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,8192,1024,0.01718933383623759
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,8192,1024,0.00831822223133511
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,8192,7168,0.14228088325924343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,8192,1024,0.14199199941423205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,8192,512,0.0059004442559348206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,8192,768,0.016488000750541687
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,8192,768,0.15226844946543375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,8192,768,0.007670222057236566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,8192,256,0.005589333259397083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,8192,512,0.007271111011505127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,8192,512,0.14134489165412054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,8192,256,0.014484445254007975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,8192,128,0.00489777781897121
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,8192,256,0.15073511335584852
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,8192,128,0.014145778285132514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,8192,128,0.0069368887278768755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,8192,3072,0.01240888900227017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,8192,64,0.004580444345871608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,8192,32,0.004909333255555895
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,8192,2560,0.1483128865559896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,8192,64,0.013778666655222574
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,8192,32,0.014442667365074158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,7168,65536,0.27697332700093585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,7168,16384,0.07418666945563422
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,7168,65536,0.12699377536773682
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,7168,65536,0.4136808978186713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,8192,4096,0.020894222789340552
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,8192,512,0.015282667345470853
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,7168,16384,0.03717866539955139
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,7168,12288,0.057513780064053006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,7168,16384,0.17606488863627115
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,8192,256,0.006945778098371293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,7168,12288,0.08196889029608832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,7168,12288,0.029189334975348577
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,7168,12288,0.14865067270067003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,8192,128,0.13332977559831408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,7168,10240,0.07076711124844022
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,7168,10240,0.025218667255507574
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,7168,65536,0.5891973177591959
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,7168,8192,0.0586951110098097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,7168,8192,0.14827911059061685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,7168,7168,0.037408000893063016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,7168,8192,0.021507556239763897
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,7168,7168,0.019682667321628995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,7168,7168,0.14268267154693604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,7168,7168,0.051960001389185585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,7168,6144,0.03326933251486884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,7168,6144,0.046111110183927745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,7168,16384,0.1067848867840237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,7168,6144,0.14578578207227919
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,7168,5120,0.027847111225128174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,7168,5120,0.016356445021099515
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,7168,5120,0.14591111077202693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,7168,5120,0.0387697782781389
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,7168,4096,0.02230666743384467
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,7168,10240,0.052526222334967725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,7168,3584,0.020248000820477802
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,7168,4096,0.014344000154071383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,7168,8192,0.04064711266093784
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,7168,4096,0.14568532837761775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,7168,3072,0.017550221747822232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,7168,10240,0.14826399750179714
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,7168,3584,0.0315688881609175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,7168,3584,0.14254844188690186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,7168,2560,0.015488889482286243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,7168,3072,0.02774488925933838
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,7168,3072,0.012220444778601328
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,7168,3072,0.1462284459008111
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,7168,2048,0.013071111506885953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,7168,2560,0.025062221619817946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,7168,2560,0.011075555450386472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,7168,2048,0.02241244415442149
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,7168,1536,0.011558222273985544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,7168,2048,0.009665778113736046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,7168,6144,0.018056000272432964
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,7168,1536,0.01936533384852939
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,7168,1024,0.008970666262838576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,7168,1536,0.147161775165134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,7168,1536,0.008996444443861643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,7168,1024,0.016774222254753113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,7168,1024,0.1466933356391059
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,7168,768,0.007620444728268518
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,7168,1024,0.00795999997191959
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,7168,512,0.006604444649484422
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,7168,4096,0.03428533342149522
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,7168,768,0.015805333852767944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,7168,768,0.007625777688291337
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,7168,256,0.005568888866239124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,7168,768,0.14623910850948757
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,7168,512,0.01480977733929952
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,7168,512,0.007159111400445302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,7168,3584,0.013382222089502545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,7168,128,0.004885333279768626
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,7168,256,0.014175999495718213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,7168,256,0.1453591055340237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,7168,256,0.0069066666894488865
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,7168,64,0.0041759999261962045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,7168,32,0.0042239998777707415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,7168,128,0.0069368887278768755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,7168,128,0.014096000128322177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,7168,128,0.1385520034366184
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,7168,2560,0.14554222424825033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,7168,64,0.014093332820468478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,7168,32,0.014108444253603617
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,6144,16384,0.05439822210205925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,7168,2048,0.14052266544765896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,6144,65536,0.20417243904537627
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,6144,12288,0.049997333023283214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,6144,16384,0.03586755527390374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,6144,65536,0.12275200419955784
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,6144,65536,0.3740871217515733
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,6144,16384,0.14346578386094835
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,6144,10240,0.033518221643235944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,6144,12288,0.07486044698291354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,6144,12288,0.02784622377819485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,6144,12288,0.14186755816141763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,6144,8192,0.03008355696996053
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,6144,10240,0.06388444370693631
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,6144,10240,0.024704888463020325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,6144,10240,0.14029599560631648
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,6144,7168,0.029583109749688044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,6144,8192,0.05348711212476095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,6144,8192,0.021115556359291077
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,6144,8192,0.14578578207227919
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,7168,512,0.1485191053814358
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,6144,65536,0.45183732774522567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,6144,7168,0.019319999549123976
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,6144,7168,0.15113066302405462
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,6144,5120,0.020536889632542927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,6144,6144,0.041893333196640015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,6144,6144,0.14660710758633083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,6144,6144,0.017704889178276062
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,6144,5120,0.03607555561595493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,6144,4096,0.0170462214284473
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,6144,5120,0.14141421847873265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,6144,5120,0.015847111741701763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,6144,3584,0.015432889262835184
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,6144,4096,0.031397332747777305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,6144,4096,0.013983999689420065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,6144,4096,0.14493244224124485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,6144,3072,0.01346933344999949
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,6144,3584,0.029410666889614526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,6144,16384,0.09549955526987712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,6144,3584,0.14947111076778835
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,6144,2560,0.011949333051840464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,6144,3072,0.01166133334239324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,6144,3072,0.14897244506412083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,6144,2560,0.02331999937693278
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,6144,2048,0.010679110884666443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,6144,2560,0.010709332923094431
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,6144,2048,0.020954666866196528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,6144,6144,0.024409777588314478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,6144,2048,0.14663022094302708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,6144,7168,0.04703199863433838
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,6144,1536,0.008614222208658854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,6144,2048,0.009511999785900116
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,6144,1024,0.0075137780772315125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,6144,1536,0.017839999662505258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,6144,1536,0.00867022242810991
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,6144,1536,0.14051111539204916
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,6144,768,0.00599911105301645
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,6144,1024,0.015835555063353646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,6144,1024,0.00794311116139094
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,6144,1024,0.14565511544545492
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,6144,512,0.004866666677925321
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,6144,768,0.015152000718646579
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,6144,768,0.007619555625650618
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,6144,768,0.15106399854024252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,6144,256,0.004259555704063839
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,6144,512,0.014163555370436775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,6144,512,0.13909333282046848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,6144,512,0.007257777783605788
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,6144,128,0.0038373333712418876
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,6144,256,0.01388711151149538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,6144,256,0.0069511110583941145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,6144,256,0.14797777599758571
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,6144,64,0.0035537779331207275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,6144,128,0.013732444081041547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,6144,32,0.00351200004418691
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,6144,128,0.1397964424557156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,6144,2560,0.14614666832817927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,6144,128,0.006909333169460297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,6144,64,0.013395555317401886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,6144,32,0.013745778136783175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,5120,65536,0.19401066833072236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,5120,16384,0.05388177765740288
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,5120,65536,0.12087021933661567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,5120,65536,0.3589484426710341
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,5120,16384,0.0942408906088935
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,5120,12288,0.04216000106599596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,5120,16384,0.14318754937913683
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,5120,16384,0.034252444903055825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,5120,10240,0.035586665074030556
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,5120,12288,0.07342311408784655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,5120,12288,0.02730488942729102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,5120,12288,0.14681333965725368
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,5120,8192,0.03067022230890062
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,6144,3584,0.012938666674825879
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,5120,10240,0.06253155734803942
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,5120,10240,0.023794665932655334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,6144,3072,0.02569955587387085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,5120,7168,0.02613688839806451
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,5120,10240,0.14727289146847194
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,5120,8192,0.051391111479865186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,5120,8192,0.02036888897418976
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,5120,6144,0.02256533337963952
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,5120,8192,0.14278577433692083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,5120,5120,0.020096888144810993
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,5120,7168,0.04639022217856514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,5120,7168,0.018717333674430847
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,5120,7168,0.14619110690222845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,5120,6144,0.017123555143674214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,5120,6144,0.1390506691402859
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,5120,4096,0.016930666234758165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,5120,5120,0.0355386667781406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,5120,5120,0.015392000476519266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,5120,5120,0.15148266156514487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,5120,4096,0.031010667483011883
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,5120,4096,0.01313866674900055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,5120,4096,0.14055466651916504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,5120,3072,0.014141332772043018
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,5120,3584,0.02880355715751648
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,5120,3584,0.15092266930474177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,5120,3584,0.012171555724408893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,5120,2560,0.012747555143303342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,5120,3072,0.14454578028784856
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,5120,2560,0.02269688910908169
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,5120,2560,0.14984799755944148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,5120,2560,0.010048888623714447
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,5120,2048,0.020311999652120803
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,5120,2048,0.14324711428748235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,5120,65536,0.4435102144877116
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,5120,2048,0.009017777939637503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,5120,1536,0.009322666459613377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,5120,1536,0.01780177818404304
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,5120,6144,0.04068355427847968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,5120,1024,0.007161777880456712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,5120,1536,0.00830577810605367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,5120,1024,0.015811555915408664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,5120,1024,0.1441804435518053
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,5120,1024,0.007983999947706858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,5120,768,0.0059004442559348206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,5120,3072,0.011058666639857821
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,5120,768,0.014819555812411837
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,5120,768,0.14476977454291448
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,5120,2048,0.010999110837777456
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,5120,512,0.005232000102599462
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,5120,512,0.014149333039919535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,5120,512,0.007285333342022366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,5120,512,0.14784088399675158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,5120,256,0.013436444103717804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,5120,256,0.1474799977408515
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,5120,128,0.0038924444880750445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,5120,256,0.006665777828958299
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,5120,1536,0.14145421981811523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,5120,128,0.013080000049538083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,5120,64,0.0035084444615576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,5120,3584,0.01534577707449595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,5120,32,0.0038293335172865125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,5120,64,0.013071111506885953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,5120,32,0.013435555828942193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,5120,3072,0.025260445144441392
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,4096,65536,0.15168000592125788
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,5120,768,0.007279111279381647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,4096,16384,0.04197600152757433
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,4096,65536,0.31893687778049046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,5120,256,0.004360888981156879
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,4096,65536,0.07680266433291964
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,4096,12288,0.03488711184925503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,4096,16384,0.0841004451115926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,4096,16384,0.02297244469324748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,4096,12288,0.06509866979387072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,4096,16384,0.14361245102352566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,4096,10240,0.029602666695912678
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,4096,12288,0.018479110466109384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,4096,12288,0.14452977975209555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,4096,8192,0.02379911144574483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,4096,65536,0.4405537711249457
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,4096,10240,0.016843555702103507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,4096,10240,0.14242133829328749
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,4096,7168,0.021299555897712708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,5120,128,0.13951555887858072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,4096,8192,0.14548356003231472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,5120,128,0.006917333437336816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,4096,6144,0.018224888377719455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,4096,7168,0.04071111149258084
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,4096,7168,0.01367022262679206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,4096,7168,0.15103556050194636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,4096,5120,0.01569422251648373
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,4096,6144,0.036850667662090726
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,4096,6144,0.012698666916953193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,4096,6144,0.14435556199815539
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,4096,4096,0.013707555830478668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,4096,5120,0.03218222326702542
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,4096,5120,0.011365333365069495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,4096,3584,0.012316444681750404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,4096,4096,0.028241776757770117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,4096,4096,0.009677333136399588
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,4096,4096,0.14278311199612087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,4096,3072,0.011037333144081963
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,4096,3584,0.02645155621899499
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,4096,3584,0.009644444617960189
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,4096,3072,0.022993778189023335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,4096,2560,0.010201777848932479
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,4096,3072,0.008960000342792934
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,4096,2560,0.020658666888872784
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,4096,2560,0.14981777138180202
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,4096,10240,0.054823110500971474
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,4096,2048,0.00865955568022198
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,4096,8192,0.04538044333457947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,4096,2560,0.00847822262181176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,4096,8192,0.014750222365061441
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,4096,1536,0.007124444676770105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,4096,2048,0.018550222118695576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,4096,2048,0.00793422261873881
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,4096,1024,0.005624889085690181
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,4096,1536,0.01646133263905843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,4096,1536,0.14172977871365017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,4096,1536,0.007484444313579136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,4096,768,0.00518844442235099
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,4096,1024,0.015127999915017022
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,4096,1024,0.0069155556460221606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,4096,1024,0.14023644394344753
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,4096,512,0.004505777938498391
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,4096,768,0.014153778553009033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,4096,768,0.006632888896597757
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,4096,768,0.14453689257303873
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,4096,256,0.004221333397759331
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,4096,512,0.01383733335468504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,4096,512,0.14503555827670628
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,4096,3584,0.1454062196943495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,4096,512,0.006400888992680444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,4096,128,0.00387999994887246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,4096,256,0.013077333569526672
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,4096,3072,0.13996977276272243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,4096,256,0.13956088489956325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,4096,256,0.006117333140638139
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,4096,64,0.0038364442686239877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,4096,32,0.003834666477309333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,4096,128,0.01293955577744378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,4096,128,0.006234666539563074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,4096,64,0.012771555946932899
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,4096,32,0.012774222426944308
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3584,65536,0.1472408903969659
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3584,16384,0.04106666644414266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,4096,2048,0.1502106719546848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,3584,65536,0.07453155517578125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3584,65536,0.3157422277662489
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3584,12288,0.03180266751183404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3584,16384,0.0836835569805569
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,3584,16384,0.0224924451775021
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,3584,16384,0.15029422442118326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,4096,5120,0.14674845006730822
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3584,10240,0.027116444375779893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,3584,12288,0.018452444010310702
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3584,12288,0.06414310799704658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,3584,12288,0.1426524453692966
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3584,8192,0.021820444199774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,3584,65536,0.4386871125962999
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,3584,10240,0.016531555189026725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,3584,10240,0.14999732706281874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3584,8192,0.04455288913514879
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,3584,8192,0.014311111635631986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3584,7168,0.04043733411365085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3584,6144,0.017484444710943434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,3584,7168,0.013408000270525614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,4096,128,0.13522044817606607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3584,6144,0.035798221826553345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,3584,6144,0.01238933371173011
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,3584,6144,0.14488089084625244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3584,5120,0.015475556254386902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3584,5120,0.031952887773513794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,3584,5120,0.1437600056330363
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3584,4096,0.013179555535316467
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3584,4096,0.027850667635599773
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,3584,4096,0.14338221814897326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3584,3584,0.012127111355463663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,3584,4096,0.009834667046864828
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3584,10240,0.05389777819315592
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3584,3584,0.02610133257177141
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3584,3072,0.010747555229398938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3584,7168,0.019816888703240287
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,3584,3584,0.009291555318567488
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,3584,8192,0.14170132742987737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3584,2560,0.009639110830095079
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3584,3072,0.022642667094866436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,3584,3072,0.008715555899673039
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,3584,3072,0.13866133160061306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3584,2048,0.00831733312871721
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3584,2560,0.020217777954207528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,3584,2560,0.008298666940795051
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,3584,2560,0.14218666818406847
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3584,1536,0.0069208890199661255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3584,2048,0.01850400037235684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,3584,2048,0.007819555699825287
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3584,1536,0.016315556234783597
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,3584,1536,0.007300444775157505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,3584,1536,0.14732444286346436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,3584,5120,0.01071999967098236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3584,768,0.004888888862397936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3584,1024,0.014820444915029736
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,3584,1024,0.006578666468461354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3584,768,0.014129777749379476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,3584,768,0.006589333216349284
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3584,512,0.004541333350870344
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,3584,768,0.14287466473049587
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,3584,512,0.006233777850866318
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,3584,512,0.14978755844963923
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3584,256,0.0038408889538711975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,3584,3584,0.14224710729387072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3584,256,0.012875555290116204
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,3584,256,0.1443271107143826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,3584,256,0.005947555518812603
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,3584,7168,0.14316622416178384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3584,128,0.012753778033786349
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,3584,128,0.005943999936183293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,3584,128,0.13765954971313477
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3584,32,0.0035048888789282907
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,3584,2048,0.14583999580807155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3584,64,0.012705778082211813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3584,1024,0.005533333453867171
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3584,32,0.0127368892232577
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,3584,1024,0.1392702261606852
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3072,16384,0.0351128876209259
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,3072,65536,0.07303910785251193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3072,65536,0.2800382243262397
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3072,16384,0.07083911365932889
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3584,512,0.013426666458447775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3072,12288,0.02608799934387207
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,3072,16384,0.022152000003390845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,3072,65536,0.43044977717929417
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3072,12288,0.05648889144261678
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,3072,12288,0.14214577939775255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3584,128,0.0035253332720862497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,3072,12288,0.018198221921920776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3072,10240,0.04822577701674568
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,3072,10240,0.14901422129737005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3584,64,0.00351200004418691
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,3072,10240,0.016361777981122334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,3072,8192,0.014128888646761576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,3072,8192,0.1435368855794271
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3072,7168,0.015607110328144498
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3072,7168,0.03626044591267904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,3072,7168,0.14128711488511828
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3072,6144,0.01404088901148902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,3072,7168,0.013008000122176277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3072,6144,0.03219733304447598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,3072,16384,0.14443200164371067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,3072,6144,0.14163733853234187
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3072,5120,0.012356444365448423
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,3072,6144,0.012127111355463663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3072,5120,0.028932445579104956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3072,10240,0.02086044516828325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,3072,5120,0.14409066571129694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,3072,5120,0.01072799993885888
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3072,4096,0.02534577747186025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,3072,4096,0.14489600393507215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,3072,4096,0.009357333183288574
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3072,8192,0.03972622089915805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,3072,3584,0.14236711131201849
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3072,3072,0.009295110901196798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,3072,3584,0.008994667066468133
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3072,3072,0.020744888318909537
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,3072,3072,0.1517146693335639
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3072,2560,0.008471999731328752
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,3072,3072,0.008608889248636033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3072,2560,0.01888088881969452
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,3072,2560,0.14135732915666369
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3072,2048,0.007248889240953658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,3072,2560,0.008198222352398766
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3072,4096,0.011009777585665384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3072,2048,0.017159110969967313
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3072,1536,0.006253333141406377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,3072,2048,0.007611555357774098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,3072,2048,0.15043377876281738
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3072,8192,0.018048000004556444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3072,1024,0.00490399988161193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3072,3584,0.00999644481473499
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3072,1536,0.015408888459205627
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,3072,1536,0.007292444507280986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,3072,1536,0.14049332671695286
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3072,768,0.004563555535342959
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3072,1024,0.014139556222491793
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,3072,1024,0.006582222051090664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,3072,1024,0.15020178423987493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3072,65536,0.12806044684516057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3072,768,0.013765333427323235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,3072,768,0.006568000134494569
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,3072,768,0.13963733779059515
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3072,512,0.013077333569526672
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,3072,512,0.006266666783226862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,3072,512,0.14986222320132786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3072,128,0.0031893334041039148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3072,256,0.012773333324326409
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,3072,256,0.00590133335855272
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3072,64,0.0031679999083280563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3072,128,0.012128888732857175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,3072,128,0.00590844452381134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3072,32,0.003175111073586676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3072,64,0.0122417774465349
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3072,32,0.012735111018021902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2560,65536,0.12055822213490804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,2560,65536,0.07238933112886217
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2560,65536,0.27835289637247723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2560,16384,0.035050666994518705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2560,16384,0.07227466503779094
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,3072,3584,0.023747555083698694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,2560,16384,0.15159289042154947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3072,512,0.003882666842805015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2560,12288,0.02716444432735443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,2560,65536,0.42759201261732316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,3072,256,0.003559111307064692
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,2560,16384,0.02164622147878011
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2560,10240,0.0232631117105484
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,2560,12288,0.14219733079274496
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,2560,12288,0.01794577803876665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2560,8192,0.019293333093325298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,3072,256,0.14139289326137966
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2560,10240,0.04640355706214905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,2560,10240,0.015816888875431485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,2560,10240,0.1454204453362359
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2560,7168,0.01722400055991279
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2560,8192,0.038973334762785174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,2560,8192,0.013784888717863293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2560,6144,0.015191111299726697
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2560,7168,0.035756445593304105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,2560,7168,0.012440000143316058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,2560,7168,0.14650044176313612
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2560,5120,0.013623110949993134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2560,6144,0.032109333409203425
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,2560,6144,0.011400000088744693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,2560,6144,0.14474844932556152
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2560,5120,0.028720888826582167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2560,4096,0.010309333602587381
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,2560,5120,0.14253955417209202
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,2560,5120,0.010664888554149203
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2560,3584,0.009729777773221334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,2560,4096,0.00943022221326828
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2560,4096,0.024844444460339014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2560,3584,0.02324177821477254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,2560,3584,0.008975110948085785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,2560,3584,0.14333866702185735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2560,3072,0.008988444175985124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2560,3072,0.019932443896929424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,2560,3072,0.008608000146018134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,2560,3072,0.14256978034973145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2560,2560,0.00798044436507755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2560,12288,0.05578666925430298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2560,2048,0.007169778148333232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2560,2560,0.0184515549076928
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,2560,2560,0.00795377790927887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,2560,2560,0.14179110527038574
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2560,1536,0.00601244428091579
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,3072,128,0.13896089129977757
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2560,2048,0.016843555702103507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,2560,8192,0.14113067256079778
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,2560,2048,0.00759733302725686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,2560,2048,0.1423128843307495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2560,1024,0.005247999810510212
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2560,1536,0.015296889675988091
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,2560,1536,0.007298666569921706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2560,768,0.004865777575307422
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,2560,1024,0.006608888920810487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2560,1024,0.01407911048995124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,2560,1024,0.14186399512820774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2560,512,0.0041884444653987885
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2560,768,0.01348000019788742
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,2560,768,0.14279733763800725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2560,256,0.003899555653333664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2560,512,0.01311644415060679
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,2560,512,0.14650221665700278
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,2560,512,0.006276444428496891
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,2560,4096,0.14308888382381862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2560,128,0.003551111039188173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2560,256,0.012450666891203986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,2560,256,0.005897777775923411
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2560,64,0.0035084444615576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2560,128,0.012447111308574677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2560,32,0.0037093332244290244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,2560,128,0.14148355854882136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,2560,128,0.005953777581453323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2560,64,0.012101333174440594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2048,65536,0.08162044154273139
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2560,32,0.012431110772821637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2048,16384,0.025789333714379206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2048,65536,0.2696044445037842
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,2048,65536,0.07117511166466607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2048,16384,0.07021955649058025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2048,12288,0.020724443925751578
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,2048,16384,0.021394666698243883
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,2048,16384,0.1435706615447998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,2560,1536,0.1456044382519192
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2048,10240,0.01795644395881229
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,2048,65536,0.4243004322052002
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2048,12288,0.05366488960054186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,2048,12288,0.15000443988376194
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2048,8192,0.01608888970481025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2048,10240,0.04573066698180305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,2048,10240,0.01552088889810774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,2048,10240,0.14667644765641955
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2048,8192,0.03834133346875509
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,2048,8192,0.013078221844302284
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,2560,256,0.14223378234439427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,2048,8192,0.14548622237311468
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2048,6144,0.012609778179062737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2048,7168,0.03533333208825853
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,2048,7168,0.012432888978057437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,2048,7168,0.14666133456759983
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2048,5120,0.01126844435930252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2048,6144,0.031023999055226643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,2048,6144,0.1487493382559882
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,2048,6144,0.011235555840863122
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2048,4096,0.01032977799574534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2048,5120,0.0281333327293396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,2048,5120,0.01034311122364468
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,2048,5120,0.14614666832817927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2048,3584,0.009223110973834991
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2048,4096,0.024519110719362896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,2048,4096,0.1497911082373725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,2048,4096,0.009308444129096137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2048,3072,0.008599999878141616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2048,3584,0.02294755478700002
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,2048,3584,0.008967111508051554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,2048,3584,0.14166755146450466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2048,2560,0.007627555893527136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2048,3072,0.019831111033757527
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,2048,3072,0.008649778034951951
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,2048,3072,0.14989777406056723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,2048,12288,0.017616000440385606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2048,2048,0.006599999964237213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2048,2560,0.018230222993426852
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,2560,768,0.006335111127959357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,2048,2560,0.007939555578761632
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2048,1536,0.0058808889653947614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2048,2048,0.01679111189312405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,2048,2048,0.1428826649983724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2048,1024,0.004864000197913912
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2048,7168,0.013991110854678683
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,2048,1536,0.007246221933099959
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,2048,1536,0.1400560008154975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2048,768,0.004523555437723796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2048,1024,0.014093332820468478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,2048,1024,0.14673066139221191
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2048,768,0.013516444298956128
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,2048,768,0.1420497761832343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2048,512,0.004121777911980947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2048,512,0.01312800000111262
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,2048,512,0.143175999323527
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2048,256,0.0034871109657817413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,2048,512,0.006193777753247156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2048,256,0.012727110750145383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,2048,256,0.14325333966149226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2048,128,0.00349866681628757
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,2048,256,0.005877333382765452
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,2048,2560,0.14137155479855007
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2048,128,0.012092444631788464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,2048,2048,0.0075866662793689305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2048,64,0.0034933334423436057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,2048,128,0.005940444353553984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,2048,32,0.0035208890007601846
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2048,1536,0.01516888870133294
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2048,64,0.012245333029164208
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1536,65536,0.07684622208277385
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,2048,32,0.012073777616024017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,2048,1024,0.006663110935025745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1536,16384,0.02427644365363651
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,1536,65536,0.06931200292375353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1536,65536,0.23099377420213488
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,2048,768,0.006540444576077991
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1536,12288,0.018605333235528734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1536,16384,0.060192002190483936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,1536,16384,0.14270044697655573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,1536,16384,0.021200888686709937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1536,10240,0.01741066740618812
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1536,12288,0.04412088791529337
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,1536,12288,0.016879111528396606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,1536,65536,0.4218444559309218
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1536,10240,0.038776887787712946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1536,8192,0.01549600064754486
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,1536,10240,0.14267199569278294
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,1536,10240,0.014727999766667684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1536,7168,0.013389333254761167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,1536,8192,0.012905778156386482
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1536,8192,0.03343288766013251
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,1536,8192,0.1425200038486057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1536,7168,0.03084889054298401
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1536,6144,0.012420444852775998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,1536,7168,0.1426346699396769
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1536,5120,0.010766222245163388
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1536,6144,0.02741066614786784
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,1536,6144,0.011054221954610614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,1536,6144,0.1463893387052748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,2048,128,0.1341093381245931
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1536,4096,0.009640889035330879
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1536,5120,0.024325332707828943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,1536,5120,0.010326222413116032
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1536,3584,0.008477333519193862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1536,4096,0.021329777108298406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,1536,4096,0.009350222018029954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,1536,4096,0.14721333980560303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1536,3072,0.007753777835104201
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1536,3584,0.020005333754751418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,1536,3584,0.008784000244405534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,1536,3584,0.14555022451612684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1536,2560,0.007229333122571309
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1536,3072,0.017900443739361234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,1536,3072,0.008332444561852349
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,1536,3072,0.14221066898769802
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1536,2048,0.005936000082227919
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1536,2560,0.016882666283183627
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,1536,2560,0.00795999997191959
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,1536,2560,0.14102399349212646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1536,1536,0.005250666704442766
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1536,2048,0.015511110424995422
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,1536,12288,0.145325329568651
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,1536,2048,0.14110222127702501
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,1536,2048,0.007639110916190677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1536,1024,0.004232888834344016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1536,1536,0.014135999812020196
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,1536,1536,0.007287999822033777
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,1536,1536,0.1478195587793986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1536,768,0.0041644444896115195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1536,1024,0.013461333182122974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,1536,1024,0.14719911416371664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1536,512,0.0038328886859946777
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1536,768,0.013063111239009433
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,1536,768,0.14510844813452825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1536,256,0.0032133333798911837
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,1536,7168,0.01183999992079205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1536,512,0.012758221891191272
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,1536,512,0.006230222268237009
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,1536,512,0.14264355765448675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1536,128,0.003152888889114062
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1536,256,0.012083555261294046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,1536,256,0.150438215997484
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,1536,256,0.005924444645643234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,1536,5120,0.14398133754730225
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1536,64,0.0028595555987623003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1536,128,0.012044444680213928
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1536,32,0.002850666642189026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,1536,128,0.005929777605666055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1536,64,0.011744889120260874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1024,65536,0.055941330062018506
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1536,32,0.01187111106183794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1024,16384,0.01980888843536377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,1024,65536,0.0675555533832974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1024,65536,0.22739201121860078
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1024,16384,0.058210664325290255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1024,12288,0.016511110795868766
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,1024,16384,0.14433599842919245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,1024,16384,0.02016888890001509
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,1024,65536,0.4185946782430013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1024,10240,0.015222221612930298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1024,12288,0.04428177740838793
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,1536,1024,0.006625777731339137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,1024,12288,0.015871110889646742
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,1024,12288,0.14431199762556288
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1024,8192,0.013826666606797112
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,1024,10240,0.014408888088332282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,1024,10240,0.14011732737223306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1024,8192,0.03233155608177185
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,1024,8192,0.14927377965715197
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,1024,8192,0.012432888978057437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1024,7168,0.029843555556403265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,1024,7168,0.011748444702890186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,1024,7168,0.14800711472829184
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1024,6144,0.01089422239197625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,1536,128,0.1382568942175971
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1024,5120,0.009974222216341231
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1024,6144,0.026664000418451097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,1024,6144,0.011083555718262991
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1024,5120,0.023599111371570166
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1024,4096,0.00833511104186376
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,1024,5120,0.14113244745466444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,1024,5120,0.010303999814722273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1024,3584,0.00775111135509279
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1024,4096,0.020636444290479023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,1024,4096,0.1454426712459988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,1024,4096,0.009215999808576372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1024,3072,0.006948444578382704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1024,3584,0.01966399947802226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,1024,3584,0.009019555317031013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,1024,3584,0.14697599411010742
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1024,2560,0.006552000012662675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1024,3072,0.01755644381046295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,1536,768,0.00629066675901413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,1024,3072,0.14190667205386692
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,1024,3072,0.00833155545923445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1024,10240,0.03806489043765598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1024,2560,0.01625866691271464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,1024,2560,0.007983110845088959
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1024,7168,0.012413333687517377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,1024,2560,0.14646044042375353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1024,2048,0.015496888094478183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,1024,2048,0.14289777808719212
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,1024,2048,0.007330666813585494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1024,1024,0.004517333375083076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1024,1536,0.014130666851997375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,1024,1536,0.14662044578128391
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1024,768,0.0038951109680864546
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1024,1024,0.013078221844302284
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,1024,1024,0.006616888774765863
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,1024,1024,0.1424151129192776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1024,512,0.0038515557017591265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,1024,6144,0.14944444762335882
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1024,768,0.012719111310111152
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,1024,768,0.00628977765639623
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,1024,768,0.1461635563108656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1024,256,0.003546666767862108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1024,512,0.01242844429281023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,1024,512,0.1489973333146837
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1024,256,0.012040889097584618
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1024,128,0.00318755561278926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,1024,256,0.14059110482533774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1024,128,0.011720000041855706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,1024,128,0.142048888736301
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,1024,128,0.005879111174080107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1024,32,0.003195555466744635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1024,64,0.011732444167137146
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,1024,32,0.011750222080283694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,768,65536,0.04198933309978909
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1024,1536,0.005192000004980299
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,768,65536,0.06586755646599664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,768,65536,0.2239075501759847
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,768,16384,0.015820443630218506
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,768,16384,0.057787557442982994
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,768,65536,0.4165511131286621
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,768,16384,0.14248711533016628
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,768,12288,0.013774221969975365
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,1024,512,0.005966222120655908
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,768,16384,0.019654222660594516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,768,12288,0.04381777842839559
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,768,12288,0.1517004436916775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,1024,256,0.005912000106440649
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,768,12288,0.015564444992277356
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1024,64,0.003198222153716617
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,768,10240,0.037323554356892906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,768,10240,0.1417306661605835
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,768,8192,0.013615111509958902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,1024,2048,0.005882666756709416
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,768,8192,0.012390221986505719
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,768,8192,0.03189599845144484
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,768,8192,0.14287377728356257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,768,7168,0.029742221037546795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,768,7168,0.14996443854437932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,768,6144,0.011053333679835001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,1024,1536,0.00721599989467197
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,768,6144,0.02679644525051117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,768,6144,0.14556622505187988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,768,6144,0.011056000159846412
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,768,5120,0.009769777456919352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,768,5120,0.01035911093155543
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,768,5120,0.14518933826022679
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,768,4096,0.008357333640257517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,768,10240,0.014912888407707214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,768,4096,0.020781333247820537
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,768,4096,0.15026666058434382
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,768,3584,0.007306666837798224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,768,4096,0.009015999734401703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,768,10240,0.014134221606784396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,768,3584,0.01923822197649214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,768,3072,0.006937777830494775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,768,3584,0.14495378070407444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,768,3584,0.008674666285514832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,768,2560,0.00598933340774642
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,768,7168,0.01240711079703437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,768,3072,0.017316444052590262
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,768,3072,0.008351999852392409
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,768,3072,0.13971643977695042
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,768,2048,0.005585777676767773
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,768,2560,0.016176000237464905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,768,7168,0.011704888608720569
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,768,2560,0.007978666987684038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,768,2560,0.14098755518595377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,768,1536,0.004898666507667965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,768,2048,0.015490666031837463
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,768,2048,0.1453688939412435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,768,2048,0.007557333343558841
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,768,1024,0.0041946665280395085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,768,1536,0.013786666923099093
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,768,1536,0.006979555719428592
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,768,768,0.004006222304370669
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,768,1024,0.006586666736337874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,768,1024,0.13991910881466335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,768,768,0.012915555801656512
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,768,5120,0.023387556274731953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,768,768,0.006280000011126201
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,768,512,0.012449777788586087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,768,256,0.0031600000543726813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,768,512,0.14607289102342394
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,768,256,0.012050666742854647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,768,128,0.003165333428316646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,768,256,0.005960888746711943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,768,128,0.011702222128709158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,768,128,0.13768711354997423
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,768,128,0.005909333212508096
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,768,64,0.0028275555620590844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,768,32,0.002809777855873108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,768,64,0.011693333586057028
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,768,32,0.01181244436237547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,512,65536,0.034099555677837796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,768,1536,0.1418773333231608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,768,1024,0.013099555340078143
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,512,16384,0.015141333142916361
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,768,512,0.0034871109657817413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,512,65536,0.0639831092622545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,512,65536,0.22557332780626085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,512,16384,0.05539200041029188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,768,512,0.005942222144868638
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,512,16384,0.14260533120897081
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,512,16384,0.01904533306757609
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,512,10240,0.011990221838156382
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,512,12288,0.015526221858130561
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,512,12288,0.04285155402289497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,512,12288,0.14096532927619085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,512,8192,0.010326222413116032
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,512,10240,0.036935110886891685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,512,10240,0.014135999812020196
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,512,10240,0.14185421996646458
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,512,7168,0.010102222363154093
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,512,8192,0.0322693321439955
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,512,8192,0.012437333663304647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,512,8192,0.14211822880638972
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,512,6144,0.009335111412737105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,512,7168,0.028775999943415325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,512,7168,0.14016621642642552
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,512,7168,0.011717333561844297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,512,5120,0.008330666356616551
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,512,6144,0.025736000802781846
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,512,6144,0.01069866700304879
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,512,5120,0.023002665903833177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,512,5120,0.14123200045691595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,512,4096,0.007947555846638149
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,768,768,0.14921511544121638
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,512,4096,0.02056799994574653
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,512,12288,0.013046222428480784
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,512,3584,0.007265778051482306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,512,4096,0.15062667263878718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,512,65536,0.40847110748291016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,512,4096,0.009291555318567488
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,512,3072,0.006583999842405319
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,512,3584,0.019542222221692402
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,512,3584,0.008643555144468943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,512,3584,0.14973778194851345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,512,2560,0.005936000082227919
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,512,3072,0.017665778597195942
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,512,3072,0.008301333420806462
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,512,3072,0.14406843980153403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,512,2560,0.016188444362746347
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,512,2560,0.007932444413503012
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,512,2560,0.14309155941009521
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,512,2048,0.015139554937680563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,512,1536,0.004881777697139316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,512,2048,0.0075937774446275495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,512,2048,0.14300978183746338
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,512,6144,0.1415084467993842
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,512,1024,0.0041786668201287585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,512,1536,0.013757333159446716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,512,1536,0.14314133591122097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,512,5120,0.010302222437328762
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,512,1024,0.013081777426931592
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,512,768,0.003896000070704354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,512,1024,0.006610666712125142
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,512,1024,0.14995288848876953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,512,512,0.003547555456558863
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,512,768,0.01272266689274046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,512,768,0.006255110932721033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,512,768,0.145449784066942
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,512,512,0.012103111379676394
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,768,256,0.14484088950686985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,512,512,0.006239999913507038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,512,256,0.011736888852384357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,512,128,0.0028568889117903183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,512,256,0.1451973385281033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,512,256,0.005886222339338726
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,512,2048,0.005495999836259418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,512,64,0.0028435554769304064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,512,128,0.011733333269755045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,512,32,0.002842666581273079
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,512,128,0.0058853332367208265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,512,128,0.13573955165015325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,256,65536,0.023734221855799358
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,512,64,0.011748444702890186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,512,32,0.011393778026103973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,256,16384,0.012437333663304647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,512,1536,0.006983111302057902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,256,65536,0.062112887700398765
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,256,65536,0.21709777249230278
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,256,12288,0.011081777513027191
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,256,16384,0.055013332102033824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,256,16384,0.018703111343913607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,256,16384,0.14251555336846247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,256,65536,0.41520176993476016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,256,10240,0.009690666364298927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,256,12288,0.04196177919705709
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,256,12288,0.015448888142903646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,256,12288,0.14883911609649658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,256,8192,0.008748444418112436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,256,10240,0.014105778601434497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,256,10240,0.03727555606100295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,256,10240,0.14275111092461482
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,256,7168,0.008296888735559251
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,512,512,0.14244533909691706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,256,8192,0.012225777738624148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,256,8192,0.03059200114674038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,256,6144,0.007919111185603673
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,256,7168,0.02910311023394267
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,256,7168,0.14058844248453775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,256,5120,0.007584888901975419
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,256,6144,0.025293333662880793
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,256,6144,0.14652889304690891
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,256,4096,0.007649777664078607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,256,5120,0.02296177711751726
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,256,5120,0.010011555420027839
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,256,5120,0.14452266693115234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,256,3584,0.007248889240953658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,256,4096,0.02057066725360023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,256,4096,0.008976000050703684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,256,4096,0.14139466815524632
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,256,3072,0.006269333263238271
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,256,3584,0.0195448895295461
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,256,3584,0.008636444807052612
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,256,3584,0.14675288730197483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,256,2560,0.005879999862776862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,256,3072,0.017508443858888414
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,256,3072,0.1411457724041409
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,256,2560,0.016315556234783597
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,256,2560,0.007982222570313348
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,256,2560,0.1398328940073649
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,512,256,0.0031742221779293488
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,256,2048,0.015138667490747241
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,256,2048,0.1408702267540826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,256,1536,0.004511111312442356
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,256,7168,0.01168444421556261
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,256,2048,0.007651555869314406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,256,6144,0.01074044406414032
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,256,1536,0.013789333403110504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,256,1024,0.003842666745185852
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,256,1536,0.007017778025733099
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,256,1536,0.14308622148301867
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,256,768,0.0035084444615576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,256,1024,0.01311377767059538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,256,1024,0.006632888896597757
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,256,1024,0.14093155331081814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,256,512,0.0031617778456873367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,256,768,0.012439111040698158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,256,768,0.006275555739800136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,256,768,0.14082666238149008
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,256,256,0.0032008888406885993
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,256,512,0.012368000215954252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,256,512,0.006212444355090459
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,256,512,0.14259910583496094
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,256,128,0.0028239999794297745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,256,3072,0.008044444852405125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,256,256,0.01204533295498954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,256,256,0.0058986664646201665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,256,256,0.14136621687147352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,256,64,0.00254133323000537
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,256,128,0.011726222104496427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,256,32,0.0028515555378463534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,256,2048,0.005192888693677055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,256,128,0.005930666708283954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,256,128,0.13603022363450792
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,128,65536,0.01972888906796773
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,256,64,0.011739555332395764
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,128,16384,0.009322666459613377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,256,32,0.011423110961914062
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,128,65536,0.05946222278806898
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,256,8192,0.15204889244503444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,128,12288,0.008616000413894653
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,128,16384,0.05464622378349304
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,128,16384,0.14172621568044028
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,128,65536,0.21988533602820504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,128,10240,0.009000000026490953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,128,12288,0.015455999308162265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,128,12288,0.04233155647913615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,128,12288,0.1444444391462538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,128,10240,0.035699556271235146
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,128,8192,0.008302222523424361
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,128,10240,0.14191910955641004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,128,10240,0.014075555735164218
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,128,8192,0.0303439994653066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,128,8192,0.012367111113336353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,128,8192,0.14517777495914036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,128,6144,0.007584888901975419
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,128,7168,0.028268443213568792
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,128,7168,0.011388444238238864
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,128,7168,0.14634844991895887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,128,6144,0.025382222400771245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,128,6144,0.010710222025712332
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,128,6144,0.14583644602033827
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,128,4096,0.007432888779375289
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,128,5120,0.022907555103302002
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,128,5120,0.010072888599501716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,128,5120,0.14517688751220703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,128,3584,0.0069644442862934535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,128,4096,0.009002666506502364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,128,3584,0.019387554791238572
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,128,3584,0.1402160061730279
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,128,16384,0.018586667047606576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,128,3072,0.006789333290523953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,128,3584,0.008635555704434713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,128,65536,0.41056889957851833
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,128,3072,0.017343999611006845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,128,3072,0.008314666648705801
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,128,3072,0.14198488659328884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,128,7168,0.007999999655617608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,128,2048,0.004899555610285865
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,128,2560,0.016137777103318107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,128,2560,0.007821333077218797
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,128,2560,0.14512177308400473
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,128,1536,0.004524444540341695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,128,2048,0.00740799970097012
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,128,5120,0.007995555798212687
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,128,1536,0.014006222287813822
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,128,1536,0.006950221955776215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,128,1536,0.14064799414740667
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,128,1024,0.01241422196229299
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,128,4096,0.020618667205174763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,128,1024,0.1385591162575616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,128,4096,0.14550222290886775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,128,768,0.003536888708670934
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,128,1024,0.006626666833957036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,128,768,0.012631999949614206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,128,512,0.003485333174467087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,128,768,0.006447111152940326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,128,768,0.14595911237928602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,128,256,0.0028604444944196274
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,128,512,0.006275555739800136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,128,512,0.15213244491153294
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,128,2560,0.005578666511509154
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,128,256,0.011755555868148804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,128,128,0.0028088889602157804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,128,256,0.14709066020117864
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,128,256,0.00589155571328269
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,128,128,0.011577777564525604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,128,64,0.002530666689078013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,384,128,128,0.005922666854328579
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,128,2048,0.015232889188660515
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,128,32,0.0026382222357723448
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,64,65536,0.014871999621391296
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,128,64,0.011732444167137146
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,64,16384,0.009722666607962714
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,128,32,0.011395555403497485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,64,12288,0.00832266691658232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,64,16384,0.05304888884226481
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,64,10240,0.009013333254390294
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,64,12288,0.041117333703570895
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,64,8192,0.008104888929261101
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,64,10240,0.035360889302359685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,64,7168,0.007859555383523306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,64,8192,0.03050044510099623
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,128,512,0.012153777811262341
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,64,6144,0.007293333609898885
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,64,5120,0.0069511110583941145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,64,7168,0.028082665469911363
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,64,4096,0.006298666612969504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,64,6144,0.02518044412136078
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,64,3584,0.006606222026877933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,64,5120,0.022661333282788593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,64,3072,0.005951111101441913
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,64,4096,0.020280000236299302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,64,2560,0.0063528890411059065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,64,3584,0.019260444574885897
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,64,2048,0.005949333310127258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,64,3072,0.01756622228357527
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,64,1536,0.004519999855094486
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,64,2560,0.015980444020695157
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,64,1024,0.0038897775941424896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,128,1024,0.0038328886859946777
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,64,768,0.003489777859714296
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,64,1536,0.013792888985739814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,64,512,0.0033173333439562055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,64,1024,0.012773333324326409
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,64,256,0.0030817778574095834
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,64,65536,0.22425333658854166
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,64,128,0.0028008888992998335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,64,512,0.012426666915416718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,64,64,0.002834666727317704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,64,256,0.011738667057620155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,64,32,0.002532444480392668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,64,128,0.011745777395036487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,64,64,0.011799111134476133
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,64,32,0.011375111010339526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,32,16384,0.009080000221729279
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,32,16384,0.052391111850738525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,32,12288,0.04042844308747186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,128,2048,0.1429697804980808
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,32,10240,0.007954667011896769
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,32,8192,0.00794666674402025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,64,2048,0.015096889601813423
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,32,7168,0.007972444097201029
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,32,8192,0.030437333716286555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,32,6144,0.007287999822033777
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,64,768,0.01273333364062839
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,32,7168,0.02814488940768772
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,32,5120,0.007121777368916406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,32,4096,0.006950221955776215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,32,6144,0.02508977717823452
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,32,5120,0.023023999399609033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,32,65536,0.016823111308945548
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,32,3072,0.0064693333374129395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,32,3584,0.019211555520693462
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,32,2560,0.006288888967699475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,32,3072,0.017535999417304993
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,32,2048,0.006230222268237009
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,32,12288,0.00831200016869439
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,32,1536,0.004442666553788715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,32,2048,0.015097777048746744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,32,1024,0.0038168889780839286
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,32,1536,0.013439999686347114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,32,768,0.0035333335399627686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,32,1024,0.013066666821638743
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,32,10240,0.03561777869860331
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,32,512,0.003141333245568805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,32,768,0.012081777883900536
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,32,256,0.0030746666921509635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,32,128,0.002854222224818336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,32,512,0.012109333442317115
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,32,256,0.011386666860845355
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,32,64,0.0024951110697454875
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,32,32,0.0028026666906144884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,32,64,0.011077333655622272
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,32,32,0.011373333632946014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,32,4096,0.020238222347365487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,32,2560,0.016132444143295288
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,65536,16384,0.37863556543986004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,16384,0.14533689286973742
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,65536,12288,0.3763004408942328
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,32,128,0.011745777395036487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,384,32,3584,0.006588444527652528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,65536,12288,0.2818666564093696
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,384,32,65536,0.22397955258687338
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,12288,0.11184710926479763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,65536,16384,0.49397956000434023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,16384,0.35983022054036456
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,65536,10240,0.23521600829230416
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,65536,8192,0.2229511208004422
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,10240,0.09394311242633396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,65536,8192,0.1908764441808065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,12288,0.2821902169121636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,65536,7168,0.17857244279649523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,8192,0.07641244596905179
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,65536,7168,0.1730124420589871
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,10240,0.24241956075032553
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,65536,10240,0.2477511035071479
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,65536,6144,0.1541902224222819
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,7168,0.06791644626193576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,65536,6144,0.14879111448923746
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,65536,5120,0.13160799609290227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,8192,0.20156710677676729
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,65536,5120,0.12445688247680664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,7168,0.18072799841562906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,65536,4096,0.1186257733239068
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,5120,0.05065155691570706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,65536,4096,0.1019573344124688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,6144,0.16103466351826987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,65536,3584,0.094376888540056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,4096,0.04246222310596042
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,5120,0.14764889081319174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,65536,3584,0.09007110860612656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,65536,3072,0.08257066541247897
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,3584,0.03766489028930664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,4096,0.14461421966552734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,65536,3072,0.07960089047749837
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,65536,2560,0.06825421916113959
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,3072,0.033724443780051336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,65536,2048,0.05562577644983927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,3072,0.14101866881052652
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,2560,0.0299288895395067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,65536,2560,0.0700942211680942
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,2560,0.14751377370622423
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,65536,1536,0.044304887453715004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,2048,0.02569777766863505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,6144,0.06015377574496799
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,65536,2048,0.06003555324342516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,65536,1536,0.04934044347869026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,65536,1024,0.03296444482273526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,2048,0.1451173358493381
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,65536,768,0.026650667190551758
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,1536,0.14435022407107884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,1536,0.020664888951513503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,1024,0.01590666671593984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,1024,0.14422399467892116
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,65536,512,0.019997333486874897
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,65536,768,0.03451910946104262
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,768,0.014476444986131458
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,768,0.14368622832828098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,65536,256,0.01310755560795466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,512,0.14369600349002415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,65536,512,0.029338667790095013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,512,0.013272888958454132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,65536,256,0.026171555121739704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,65536,128,0.011696000066068439
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,256,0.15088621775309244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,65536,64,0.009334222310119206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,3584,0.14299910598331028
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,65536,128,0.025053333905008104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,128,0.011703111231327057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,65536,128,0.1359822220272488
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,65536,32,0.009720000127951304
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,65536,64,0.024352888266245525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,65536,32,0.0244862222009235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,16384,16384,0.10107644399007161
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,16384,65536,0.39210134082370335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,16384,65536,0.46150488323635525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,16384,16384,0.1220746702618069
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,65536,1024,0.040179553959104754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,65536,0.15298488405015734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,16384,12288,0.07821244663662381
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,16384,0.04360711243417528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,16384,12288,0.09285155269834731
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,16384,10240,0.0659137765566508
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,12288,0.03382933470937941
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,12288,0.15204444196489122
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,65536,256,0.012157333393891653
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,16384,10240,0.07983911037445068
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,16384,8192,0.05593422386381361
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,10240,0.02943377693494161
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,16384,7168,0.05000977714856466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,10240,0.1445982191297743
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,16384,8192,0.06659199794133504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,8192,0.14678043789333767
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,16384,7168,0.05916621949937609
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,7168,0.022476444641749065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,7168,0.1426711082458496
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,16384,6144,0.0529368884033627
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,16384,5120,0.03663466705216302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,6144,0.14460355705685085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,6144,0.02004888819323646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,65536,0.5087226761711968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,16384,5120,0.047021332714292735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,16384,4096,0.03048444456524319
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,16384,0.15909511513180202
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,5120,0.01817244456874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,16384,4096,0.03800800111558702
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,4096,0.1458915604485406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,4096,0.01551644504070282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,8192,0.024744888146718342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,3584,0.1410177813635932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,16384,3072,0.0247706671555837
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,3584,0.01421688828203413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,16384,3072,0.03227555420663621
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,3072,0.14908622370825872
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,16384,2560,0.020936000678274367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,3072,0.013177778157922955
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,16384,2560,0.028743111424975928
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,2560,0.14086933930714926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,5120,0.14422489537133112
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,2560,0.012229333321253458
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,16384,2048,0.017084444562594097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,16384,2048,0.025137777129809063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,16384,3584,0.027154665854242113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,2048,0.14781778388553196
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,2048,0.011111111276679568
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,1536,0.009360888765917884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,16384,1536,0.021976000732845728
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,1536,0.14891821808285183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,16384,1024,0.011063999599880643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,16384,3584,0.03513600097762214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,16384,768,0.009681777821646797
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,16384,6144,0.04215111003981697
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,16384,1024,0.019184889064894784
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,1024,0.00793866647614373
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,1024,0.14811555544535318
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,16384,512,0.007996444072988298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,16384,768,0.017493334081437852
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,768,0.14855378203921848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,768,0.007638221813572778
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,16384,512,0.016335111525323655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,16384,256,0.005590222362014983
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,512,0.1417893303765191
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,512,0.007175111108356052
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,16384,128,0.004911111046870549
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,16384,256,0.015278221832381355
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,256,0.14779733286963567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,16384,64,0.00489155575633049
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,16384,128,0.13344621658325195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,128,0.006918222126033571
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,16384,32,0.004927999857399198
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,16384,64,0.014459555347760519
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,16384,32,0.015107555521859063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,16384,1536,0.013845333622561561
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,12288,65536,0.2965573204888238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,12288,16384,0.07824266619152494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,12288,65536,0.39446134037441677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,12288,16384,0.1075893309381273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,65536,0.1345813274383545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,12288,12288,0.06103733513090345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,16384,0.15456177128685847
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,12288,12288,0.0819715526368883
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,12288,10240,0.05144888824886746
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,12288,0.031495110856162176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,16384,256,0.006907555378145642
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,16384,128,0.01477866702609592
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,12288,0.14190667205386692
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,12288,8192,0.04489244355095757
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,12288,10240,0.07059644328223334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,10240,0.027089777919981215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,12288,8192,0.05890666776233249
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,10240,0.1421911054187351
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,8192,0.02273333403799269
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,8192,0.14537778165605333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,12288,7168,0.05264888869391548
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,7168,0.02055111030737559
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,7168,0.14315556155310735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,65536,0.4977591302659776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,12288,6144,0.04822133315934075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,12288,5120,0.029870222012201946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,6144,0.018631110588709515
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,6144,0.140794661309984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,12288,5120,0.04205866654713949
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,16384,0.040009776751200356
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,5120,0.01681866745154063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,5120,0.14972799354129368
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,12288,3584,0.021905778182877436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,12288,4096,0.03437599870893691
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,4096,0.014831110835075378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,4096,0.1434275574154324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,12288,3584,0.03130577670203315
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,12288,3072,0.018957333432303537
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,3584,0.14049955209096274
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,12288,3072,0.02892177634769016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,3072,0.012803555362754397
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,3072,0.14200354946984187
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,12288,7168,0.03772977656788296
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,12288,2560,0.025623111261261836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,2560,0.011680888632933298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,12288,2048,0.013942221800486246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,12288,6144,0.03467199868626065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,12288,2048,0.02274133265018463
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,2048,0.14327111509111193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,12288,1536,0.010767999622556897
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,2048,0.010129777921570672
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,12288,4096,0.02458577851454417
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,12288,1536,0.020564445190959506
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,12288,1024,0.009831111464235518
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,1536,0.14680443869696722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,1536,0.008622222476535374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,12288,1024,0.017901332841979135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,12288,768,0.008259555531872643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,1024,0.00794222205877304
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,3584,0.013678222066826291
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,12288,768,0.016332444217469957
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,12288,512,0.006613333192136552
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,768,0.007266666326257918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,768,0.14115111033121744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,12288,512,0.01514488955338796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,512,0.0069395556218094295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,512,0.1437022156185574
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,2560,0.1417048904630873
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,12288,256,0.014420444766680399
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,256,0.0069653333889113525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,12288,128,0.013791110780504016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,128,0.13462755415174696
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,12288,128,0.006966222491529252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,1024,0.14534933037228054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,12288,64,0.0047786666287316214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,12288,32,0.004917333523432414
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,12288,64,0.014090667168299357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,12288,32,0.014132445057233175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,12288,256,0.0052177777720822235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,10240,65536,0.22864445050557455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,12288,256,0.14059466785854763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,10240,16384,0.07150844732920329
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,10240,65536,0.3519599967532688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,10240,16384,0.0941351122326321
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,65536,0.1308871110280355
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,10240,12288,0.051186667548285596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,16384,0.15336710876888698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,16384,0.03900711072815789
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,10240,12288,0.07347110907236735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,10240,10240,0.046775112549463906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,12288,0.030001779397328694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,12288,0.14681866433885363
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,12288,2560,0.016501333978441026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,10240,8192,0.036249776681264244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,10240,0.14369155300988093
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,10240,0.026131555438041687
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,10240,8192,0.05330844389067756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,10240,7168,0.03056710958480835
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,8192,0.02193333374129401
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,8192,0.1452960040834215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,12288,128,0.004691555682155821
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,65536,0.49388710657755536
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,10240,6144,0.027991112735536363
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,10240,7168,0.047804445028305054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,10240,5120,0.02415200074513753
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,7168,0.14287732707129583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,6144,0.01793599956565433
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,10240,6144,0.04383733206325107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,6144,0.14112444718678793
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,10240,4096,0.02015733387735155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,10240,5120,0.0372995568646325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,5120,0.14286932680341932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,5120,0.016511999898486666
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,10240,3584,0.017793777916166518
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,4096,0.014563555518786112
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,4096,0.1407617727915446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,10240,4096,0.030692444907294378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,10240,3072,0.015370666980743408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,10240,3584,0.0281368891398112
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,3584,0.013311111264758639
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,10240,3072,0.026345777842733595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,10240,2560,0.013314666847387949
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,3072,0.012362666428089142
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,3072,0.14039377371470133
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,10240,2048,0.011506666739781698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,10240,10240,0.06381333536571927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,2560,0.1477644443511963
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,2560,0.010957332948843638
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,10240,1536,0.009720889230569204
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,10240,2048,0.021568889419237774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,2048,0.009574222067991892
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,384,128,128,0.14288177755143908
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,2048,0.14413244194454616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,10240,1536,0.019205333458052743
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,1536,0.14135911729600695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,1536,0.00832711077398724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,10240,768,0.0068853331936730286
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,10240,1024,0.01646577815214793
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,7168,0.019951999187469482
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,1024,0.14633156193627253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,1024,0.007657777931955125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,10240,512,0.00555644432703654
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,10240,768,0.01552177800072564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,768,0.007311111523045435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,768,0.1454026699066162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,10240,512,0.014724445011880664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,512,0.14628888501061335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,512,0.006955555743641323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,10240,256,0.013793778088357715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,256,0.14665510919358996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,256,0.006777777853939269
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,3584,0.1416462262471517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,10240,128,0.013668444421556262
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,10240,128,0.14053332805633545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,10240,64,0.0038275557259718576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,10240,32,0.0038951109680864546
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,10240,128,0.006907555378145642
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,10240,2560,0.023972445064120825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,10240,64,0.01313066648112403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,10240,32,0.01350933313369751
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,8192,65536,0.20595999558766684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,10240,1024,0.007967111137178209
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,8192,16384,0.054732445213529796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,8192,65536,0.30972443686591256
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,8192,16384,0.08397599723603989
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,65536,0.08292710781097412
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,8192,12288,0.042690667841169566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,16384,0.02641244398223029
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,16384,0.14287643962436253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,8192,12288,0.06532355811860827
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,8192,10240,0.03665333324008518
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,12288,0.020465777979956735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,12288,0.14683822790781656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,10240,128,0.0038542221817705366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,8192,8192,0.030906667311986286
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,8192,10240,0.056984888182746045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,10240,0.017755554782019723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,10240,0.1426888836754693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,8192,7168,0.027842667367723253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,8192,0.015439111325475903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,65536,0.35338489214579266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,8192,0.14598577552371556
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,8192,6144,0.02456088860829671
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,8192,7168,0.04269599914550781
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,7168,0.1392106612523397
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,7168,0.014466666513019137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,8192,5120,0.02071466710832384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,8192,6144,0.038176000118255615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,6144,0.013438222308953604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,6144,0.1478053331375122
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,8192,4096,0.018191999859280057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,8192,5120,0.032913777563307024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,5120,0.01239377756913503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,5120,0.13968177636464438
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,8192,3584,0.015646222564909194
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,8192,4096,0.028512001037597656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,4096,0.14148710833655462
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,8192,3072,0.013843555417325763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,8192,3584,0.025736000802781846
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,10240,256,0.004459555364317364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,3584,0.1459102233250936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,8192,2560,0.011931555966536203
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,8192,3072,0.023652444283167522
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,3072,0.14008355140686035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,8192,2560,0.02198222279548645
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,2560,0.008748444418112436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,8192,2048,0.010680889089902243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,2560,0.14188978407118055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,8192,8192,0.046903110212749906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,8192,1536,0.008765333228641087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,8192,2048,0.02021511064635383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,2048,0.007921777665615082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,2048,0.14685600333743626
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,8192,1024,0.007239110767841339
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,8192,1536,0.017856889300876193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,1536,0.14528621567620173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,8192,768,0.005531555662552516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,1024,0.006599999964237213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,8192,768,0.014837332897716098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,768,0.14563911490970188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,8192,512,0.004622222234805425
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,768,0.006592000110281839
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,3584,0.010025777750545079
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,8192,512,0.014025777578353882
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,512,0.1449893315633138
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,512,0.00628266649113761
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,3072,0.009017777939637503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,8192,256,0.013442666994200813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,8192,128,0.0037582222786214617
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,256,0.005968889014588461
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,256,0.14974578221638998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,8192,128,0.013094222380055321
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,128,0.005995555470387141
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,8192,64,0.003552000141806073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,128,0.1340657737520006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,8192,32,0.0038524443904558816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,1536,0.007250666618347168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,8192,64,0.01276177747382058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,8192,32,0.013234666652149625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,8192,1024,0.01570133368174235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,7168,65536,0.17903467019399008
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,7168,16384,0.05077777637375725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,7168,65536,0.30192089080810547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,65536,0.08001689116160075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,7168,12288,0.04306755463282267
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,16384,0.1449715561336941
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,16384,0.02517066730393304
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,7168,12288,0.06422577963935004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,12288,0.1471724510192871
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,7168,10240,0.03417866759830051
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,65536,0.34758755895826554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,7168,10240,0.05597244368659126
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,10240,0.01719199948840671
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,10240,0.14828533596462673
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,7168,8192,0.028312000963422988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,7168,8192,0.04532088836034139
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,8192,0.14541599485609266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,8192,4096,0.011005333728260465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,7168,16384,0.08319911029603747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,7168,7168,0.024618667032983568
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,8192,0.015071999695565967
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,7168,6144,0.0218542218208313
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,7168,7168,0.041399998797310725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,8192,256,0.004173333446184794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,7168,0.013991999957296582
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,7168,0.14601422680748835
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,7168,5120,0.018248889181349013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,7168,6144,0.03638933433426751
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,6144,0.14069422086079916
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,7168,5120,0.0325911111301846
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,7168,4096,0.015463999576038785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,12288,0.019324445062213473
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,5120,0.14540355735354954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,5120,0.012060444388124677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,7168,3584,0.013967111706733704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,7168,4096,0.0273991111252043
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,4096,0.01035733355416192
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,4096,0.14235466056399876
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,7168,3072,0.012599111431174807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,3584,0.14697333176930746
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,3584,0.00963466697269016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,7168,2560,0.011164444188276926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,7168,3072,0.02368444369898902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,8192,1024,0.14164888858795166
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,3072,0.0086666668454806
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,7168,2048,0.010275555153687796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,7168,2560,0.02131822208563487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,2560,0.14015911685095891
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,2560,0.008295999632941352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,7168,1536,0.008777778181764815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,7168,2048,0.019339554839664035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,2048,0.1469448937310113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,2048,0.00795022232664956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,7168,1024,0.006933333145247565
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,7168,1536,0.017147555947303772
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,1536,0.007184888753626082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,1536,0.14636622534857854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,7168,768,0.005606222069925732
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,7168,1024,0.015457777513398064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,1024,0.006640000061856375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,1024,0.14185778299967447
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,6144,0.013045333325862885
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,7168,768,0.014567999376191033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,768,0.14185688230726454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,768,0.006341333190600078
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,7168,256,0.0042151109211974675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,7168,512,0.013770666387346057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,512,0.14462578296661377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,512,0.006226666685607698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,7168,128,0.003846222327815162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,7168,256,0.013404444687896304
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,7168,3584,0.02535466684235467
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,256,0.14537955654992
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,7168,64,0.0035617777870761026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,7168,128,0.01311822235584259
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,128,0.005993777679072485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,7168,32,0.0038480001191298165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,128,0.1379973358578152
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,7168,3072,0.14157244894239637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,7168,64,0.012844444149070315
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,7168,32,0.012423111332787408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,6144,65536,0.16516266928778753
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,6144,16384,0.04879822333653768
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,6144,65536,0.2614053355322944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,65536,0.07675555679533216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,6144,16384,0.07373688618342082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,6144,12288,0.036325332191255354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,16384,0.14790577358669704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,16384,0.023541332946883306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,6144,10240,0.031587554348839655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,7168,512,0.00498933345079422
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,6144,12288,0.05698844459321764
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,12288,0.14004355006747776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,6144,8192,0.02606222199069129
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,6144,10240,0.04909244510862562
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,10240,0.01682222220632765
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,10240,0.14772621790568033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,65536,0.3458071019914415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,6144,7168,0.022520888182852004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,6144,8192,0.04079466727044847
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,8192,0.014487110906177096
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,8192,0.1403528849283854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,6144,7168,0.03610666592915853
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,7168,256,0.005920889063013925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,7168,0.14258221785227457
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,6144,5120,0.01757955551147461
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,6144,6144,0.03267733256022135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,6144,0.1407111088434855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,6144,0.012688000169065265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,6144,4096,0.014877332581414117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,5120,0.1470364464653863
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,5120,0.011505777637163797
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,6144,4096,0.024804444776640996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,6144,3584,0.013131555583741931
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,4096,0.009992889232105678
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,4096,0.14573066764407688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,6144,3584,0.02331200076474084
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,3584,0.009182222187519073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,3584,0.14045333862304688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,12288,0.01869066721863217
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,3072,0.14673954910702175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,6144,2560,0.010262221925788457
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,3072,0.0083155557513237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,6144,2560,0.02020711037847731
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,6144,6144,0.020058666666348774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,2560,0.1452604399787055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,6144,2048,0.009295110901196798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,7168,0.013770666387346057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,2560,0.007990222010347579
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,6144,1536,0.007967111137178209
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,6144,2048,0.018343110879262287
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,2048,0.14002222485012478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,2048,0.007611555357774098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,6144,1024,0.0058337777025169795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,6144,5120,0.029860443539089624
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,6144,1536,0.016081778539551627
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,1536,0.007272889216740926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,1536,0.14467288388146296
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,6144,768,0.005555555638339784
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,6144,1024,0.014782221780882942
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,1024,0.14192622237735325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,6144,768,0.0140364451540841
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,768,0.1386400063832601
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,768,0.006575999988449945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,6144,3072,0.011935999823941125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,6144,256,0.004214222232500712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,512,0.006224000205596288
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,512,0.13995822270711264
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,6144,3072,0.021903999977641638
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,6144,128,0.003846222327815162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,6144,256,0.01292888902955585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,256,0.005960000058015187
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,256,0.14367022779252794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,6144,128,0.012808889150619507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,6144,128,0.13979466756184897
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,128,0.005936000082227919
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,6144,32,0.0038648889296584656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,6144,64,0.01275288893116845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,6144,32,0.01276711126168569
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,5120,65536,0.14550311035580105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,5120,65536,0.2577306694454617
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,5120,16384,0.03996088769700792
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,65536,0.07417422533035278
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,6144,512,0.004537777768241035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,5120,12288,0.033253331979115806
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,5120,16384,0.07291644149356417
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,16384,0.022276444567574397
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,16384,0.1452817784415351
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,6144,512,0.013394667042626275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,5120,10240,0.027150221996837195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,12288,0.018325333793958027
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,5120,12288,0.0556817783249749
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,5120,8192,0.023361777265866596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,5120,10240,0.04811733298831516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,10240,0.14487644036610922
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,10240,0.016214221715927124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,5120,7168,0.020206222931543987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,6144,64,0.0035235554807715942
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,5120,8192,0.039515554904937744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,8192,0.014138667119873894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,5120,6144,0.017091555727852713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,5120,7168,0.0351839992735121
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,7168,0.013283555706342062
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,7168,0.14225688245561388
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,5120,5120,0.015189333094490899
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,5120,6144,0.0325164430671268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,6144,0.012255111502276527
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,6144,0.14776533179812962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,5120,4096,0.0129439996348487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,5120,5120,0.029215110672844782
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,5120,0.010967999696731567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,5120,0.14407466517554388
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,5120,3584,0.011691555380821228
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,6144,1024,0.006575111299753189
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,5120,4096,0.02446577780776554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,4096,0.14305867089165583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,5120,3072,0.011213333242469363
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,5120,3584,0.02313599983851115
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,3584,0.008660444782839881
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,3584,0.14055111673143175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,5120,2560,0.009285333255926767
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,5120,3072,0.021552888883484736
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,3072,0.008314666648705801
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,3072,0.14587910970052084
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,12288,0.1481191052330865
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,5120,2560,0.01993333299954732
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,5120,2048,0.008270222279760573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,65536,0.34062666363186306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,2560,0.007983999947706858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,5120,1536,0.0069368887278768755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,5120,2048,0.017142222987280954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,2048,0.143939561314053
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,8192,0.14567822880215114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,5120,1024,0.005561777700980504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,5120,1536,0.015488889482286243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,1536,0.0069324444565508105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,1536,0.14552888605329725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,5120,768,0.004912000149488449
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,1024,0.14681688944498697
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,1024,0.006576889091067844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,5120,768,0.013767999907334646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,5120,512,0.004211555752489301
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,768,0.006272000157170826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,768,0.14662933349609375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,4096,0.009544889132181803
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,512,0.006274666637182236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,512,0.14464888307783338
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,5120,256,0.01274311128589842
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,256,0.14565332730611166
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,5120,128,0.0034942221310403612
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,256,0.005962666538026597
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,2560,0.14294221666124132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,5120,128,0.01276088919904497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,5120,64,0.0032106666929192017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,5120,128,0.13908444510565862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,2048,0.0075528886583116316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,5120,128,0.005917333480384614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,5120,32,0.0035253332720862497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,5120,64,0.012720000412729053
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,4096,65536,0.10961955785751343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,5120,32,0.012720000412729053
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,5120,1024,0.014441778262456259
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,4096,65536,0.21954133775499132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,4096,16384,0.06147288613849216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,16384,0.14822667174869114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,5120,512,0.013448889056841532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,4096,12288,0.02680355476008521
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,5120,256,0.003863111138343811
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,4096,12288,0.048249777820375234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,65536,0.33224089940388996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,12288,0.01756266587310367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,12288,0.14640177620781794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,4096,10240,0.02214488883813222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,4096,8192,0.018519111805491976
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,4096,10240,0.04008888867166307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,10240,0.015916445189052157
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,10240,0.1424880027770996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,4096,7168,0.016846223009957206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,4096,8192,0.033999111917283796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,8192,0.013767110804716745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,8192,0.14686755339304605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,4096,7168,0.03138400117556254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,4096,6144,0.014278221461508008
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,7168,0.14701688289642334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,7168,0.0127697777416971
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,4096,16384,0.03516622384389242
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,4096,6144,0.028763555818133887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,6144,0.011420444481902652
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,6144,0.14478311273786756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,4096,4096,0.01130222198035982
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,4096,5120,0.026263111167483862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,65536,0.07218755616082086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,5120,0.13991288344065347
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,4096,3584,0.010803555448849997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,4096,4096,0.021928000781271193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,4096,0.14773333072662354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,16384,0.021372444099850122
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,4096,3584,0.02104622291194068
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,3584,0.008671110702885522
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,4096,3072,0.01903022163444095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,3072,0.1428640021218194
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,4096,2560,0.008658666577604082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,3072,0.008267555799749162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,4096,2560,0.017831110292010836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,2560,0.007713778151406183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,2560,0.14525689019097224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,4096,2048,0.007835555407736037
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,4096,2048,0.01587199999226464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,2048,0.007445333732499017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,2048,0.14424533314175075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,4096,1536,0.006565333240562015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,4096,5120,0.012795555922720166
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,4096,1536,0.014827556080288358
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,1536,0.1443679995006985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,4096,1024,0.00526577772365676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,1024,0.006610666712125142
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,4096,1024,0.013744889034165276
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,5120,0.010281778044170803
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,4096,0.008964444200197855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,4096,768,0.013431111143694984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,4096,3072,0.009670221971140968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,4096,512,0.0041848888827694785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,768,0.14165155092875162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,3584,0.14205333921644422
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,768,0.0063982220987478895
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,4096,256,0.003823111040724648
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,4096,512,0.013074666261672974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,512,0.006080889039569431
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,512,0.14143199390835232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,4096,128,0.0034968890249729156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,4096,256,0.012488000094890594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,256,0.005934222290913264
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,4096,128,0.012082666986518435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,128,0.0058986664646201665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,4096,64,0.0034968890249729156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,4096,32,0.003584889074166616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,4096,64,0.012432888978057437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,4096,32,0.012096888489193387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,4096,1536,0.006970666348934174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3584,65536,0.10179822312461005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,1024,0.14386755890316433
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3584,65536,0.21680977609422472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3584,16384,0.034099555677837796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,65536,0.0709022217326694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3584,16384,0.06025955412122938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,16384,0.021362667282422382
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3584,12288,0.028917332490285236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,16384,0.14587910970052084
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,65536,0.32602132691277397
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3584,12288,0.045928888850741915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,12288,0.017263111140992906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3584,10240,0.024365334047211543
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,12288,0.14245688915252686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3584,8192,0.02014844450685713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,128,0.13284622298346624
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3584,10240,0.03986933496263292
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,10240,0.1430799961090088
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3584,8192,0.03378399875428941
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,8192,0.013454222016864352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,8192,0.1395991113450792
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3584,7168,0.03090933296415541
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3584,6144,0.015841777125994366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,7168,0.012179555164443122
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,7168,0.14004711310068765
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,4096,768,0.004607999904288186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,6144,0.1380808883243137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,6144,0.011394666300879585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3584,5120,0.014291556345091926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3584,5120,0.02581600017017788
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,5120,0.010046222143703038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,5120,0.1391351090537177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3584,4096,0.021631111701329548
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,4096,0.13990667131212023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,4096,256,0.14062755637698704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3584,3584,0.009695111049546136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,4096,0.009053332938088311
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,10240,0.01555199921131134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3584,3584,0.020344000723626878
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,3584,0.008661333057615492
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3584,7168,0.01816444430086348
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,3584,0.14051911565992567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3584,2560,0.008270222279760573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3584,3072,0.018572444717089336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,3072,0.0083137775460879
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,3072,0.14274844858381483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3584,2560,0.01721333298418257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3584,2048,0.00720266666677263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,2560,0.14315022362603083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,2560,0.007972444097201029
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3584,1536,0.0060159998635451
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3584,2048,0.015832000308566622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,2048,0.007393777370452881
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3584,1024,0.0052435555391841466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3584,6144,0.028480887413024902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3584,1536,0.014548444085650973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,1536,0.0069466663731469055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,1536,0.14525866508483887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3584,768,0.004528000122971005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3584,1024,0.013767999907334646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,1024,0.006623111251327727
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,1024,0.14186844560835096
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3584,512,0.004257777912749184
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3584,768,0.013119999733236102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3584,4096,0.010999999940395355
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,768,0.006484444356626934
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3584,256,0.0038355555799272326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3584,512,0.013033777475357056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,512,0.00590133335855272
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,512,0.14589777257707384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3584,128,0.0035617777870761026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3584,256,0.01240711079703437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,256,0.1402319934633043
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3584,64,0.0033813334173626373
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3584,128,0.012120889292822944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,128,0.005941333456171884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,128,0.1368417739868164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3584,32,0.0035173334181308746
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3584,3072,0.009081777599122789
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3584,64,0.012039999994966718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3072,65536,0.0924453337987264
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3584,32,0.011777777638700275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3072,16384,0.03154222170511881
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3072,65536,0.21150844626956514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3072,16384,0.058507555060916476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3072,12288,0.02206755512290531
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,16384,0.14401511351267496
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,2048,0.13984710640377468
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,16384,0.020939555433061387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3072,12288,0.04459822177886963
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,12288,0.1438995599746704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,12288,0.017200888858901132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3072,10240,0.018785778019163344
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,3584,768,0.14290577835506862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3072,10240,0.03883822096718682
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3072,8192,0.015867556134859722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,3584,256,0.005903999838564131
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,8192,0.1395182212193807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3072,7168,0.014135999812020196
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,8192,0.013458666702111563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3072,7168,0.03052088949415419
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,7168,0.14275644885169134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3072,6144,0.012359110845459832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,7168,0.012083555261294046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,65536,0.06985955768161349
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3072,6144,0.028134223487642076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,6144,0.01070844464831882
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3072,5120,0.011056889262464313
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,6144,0.14186310768127441
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3072,5120,0.025384000606007043
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,65536,0.32983022265964085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,5120,0.14209689034356013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3072,4096,0.009977777798970541
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,10240,0.0155102229780621
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,4096,0.008961777720186446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,4096,0.14063199361165366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3072,3584,0.009330666727489894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3072,3584,0.019479110836982727
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3072,8192,0.03319733341534933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,3584,0.14196977350446913
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3072,3072,0.008613333106040955
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3072,3072,0.018226666582955253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,3072,0.008296888735559251
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,3072,0.1403155591752794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3072,2560,0.01686311099264357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,2560,0.14149778419070774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3072,2048,0.006476444502671559
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,2560,0.007807110746701558
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3072,2048,0.015632000234391954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3072,1536,0.0057884446448749965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,2048,0.14034667279985216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,2048,0.007307555940416124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,5120,0.010037333601050908
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,1536,0.0069164443347189165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,1536,0.14031733406914607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3072,1024,0.004560889055331548
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3072,1024,0.013752000199423896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,1024,0.006574222197135289
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,10240,0.14538399378458658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3072,768,0.004521777646409141
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,1024,0.14070755905575222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3072,512,0.003846222327815162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,3584,0.008632889224423302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3072,768,0.013391111459996967
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,768,0.1448097758822971
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3072,512,0.01279022213485506
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3072,256,0.003531555748648114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,512,0.005998222364319696
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,512,0.15075466367933485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3072,2560,0.00757422215408749
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3072,128,0.0035253332720862497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3072,256,0.012426666915416718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,256,0.14460444450378418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3072,64,0.0034711110509104202
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3072,128,0.01205244412024816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,128,0.005889777921968036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,3072,128,0.1398622194925944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2560,65536,0.07671733034981622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3072,32,0.012362666428089142
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3072,1536,0.01477777792347802
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2560,65536,0.20849688847859701
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2560,16384,0.02833599845568339
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3072,4096,0.02102222210831112
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,65536,0.06856355402204725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2560,16384,0.056808001465267606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,768,0.006286222073766921
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,16384,0.020793777373101976
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2560,12288,0.022533333963818018
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2560,12288,0.04403644469049242
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,12288,0.01684977776474423
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2560,10240,0.020255110330051847
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,3072,256,0.00591466658645206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2560,10240,0.03869511021508111
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,3072,32,0.003154666680428717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,10240,0.1426373322804769
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2560,8192,0.015142222245534262
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,3072,64,0.0121760004096561
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2560,8192,0.03256888853179084
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2560,7168,0.013536888692114087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,8192,0.012368888490729861
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,8192,0.13962222470177546
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2560,6144,0.012367111113336353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,7168,0.011743110915025076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2560,7168,0.030144890149434406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,7168,0.14689332909054226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2560,5120,0.010694222317801582
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2560,6144,0.027781334188249376
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,6144,0.01071911139620675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,6144,0.1476960049735175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2560,5120,0.025137777129809063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,5120,0.14004533820682102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,5120,0.010024000373151567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2560,4096,0.009349333743254343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,16384,0.1438488827811347
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2560,3584,0.008994667066468133
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2560,4096,0.020622221959961783
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,4096,0.14807644155290392
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,12288,0.14251111613379583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2560,3072,0.008286221987671321
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2560,3584,0.01926488843229082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2560,3072,0.018207111292415194
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,10240,0.014821334017647637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,3072,0.1442693339453803
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2560,2560,0.007599111232492659
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,3072,0.008175111479229396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2560,2560,0.016910221841600206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,2560,0.14198577404022217
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,2560,0.007657777931955125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2560,2048,0.015480000111791821
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,2048,0.14225067032708064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2560,1536,0.005899555567238066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,2048,0.007296000089910295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,65536,0.3253573311699761
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2560,1536,0.014504889647165934
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2560,1024,0.0047288888858424295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,1536,0.1417093276977539
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,4096,0.009014222357008193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,1536,0.006963555183675554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2560,768,0.004212444441186057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2560,1024,0.013543999857372709
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,1024,0.006601777755551868
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,1024,0.1422142187754313
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,3584,0.1398951080110338
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2560,512,0.0038853333228164246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2560,768,0.013223999904261695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,768,0.1410177813635932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,768,0.006274666637182236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2560,256,0.003559999995761447
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2560,512,0.01276177747382058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,512,0.005949333310127258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,512,0.14232977231343588
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2560,128,0.0034977777136696708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2560,256,0.01240888900227017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,256,0.00589155571328269
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,256,0.14065689510769314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2560,128,0.012089778151777057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2560,2048,0.006266666783226862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,2560,128,0.1324924495485094
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2560,32,0.0031991110493739447
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2560,64,0.012044444680213928
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2048,65536,0.06013244390487671
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2560,32,0.012368000215954252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2048,16384,0.022626666559113398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2048,65536,0.17011822594536674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,65536,0.06752533382839628
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2048,16384,0.046966221597459584
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,16384,0.020218667056825426
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,16384,0.14276977380116782
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,65536,0.3272328906589084
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2048,10240,0.016791999340057373
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,12288,0.0162435554795795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2048,12288,0.036489778094821505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,3584,0.008662222160233391
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2048,8192,0.014492443866199918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,10240,0.01386044422785441
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2048,10240,0.032587554719712995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2048,8192,0.02797777785195245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,8192,0.012303111453851065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,8192,0.13996977276272243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2048,7168,0.01311644415060679
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,2560,128,0.005935110979610019
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2048,6144,0.011607111328177981
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2048,7168,0.025683555338117812
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,7168,0.011459555890825061
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,7168,0.14079822434319392
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2048,5120,0.010602666272057427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2048,6144,0.023956444528367784
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,6144,0.0107004443804423
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,6144,0.1434897714191013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2048,4096,0.009005332986513773
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,5120,0.010016889207892949
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,5120,0.1473662190967136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2048,12288,0.020767110917303298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2048,3584,0.00851644410027398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,4096,0.008985777695973715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,4096,0.14813866880204943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,12288,0.14222310649024114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,3584,0.00868800034125646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2048,3072,0.007631111476156447
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,3584,0.14683377742767334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,10240,0.14339999357859293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,3072,0.008031999899281396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2048,3072,0.0162462227874332
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,3072,0.14178666803571913
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2048,2560,0.015522667103343539
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2560,64,0.0031840000301599503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,2560,0.14870577388339573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2048,2048,0.00562755556570159
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2048,2048,0.014152000347773233
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,2048,0.14914933840433756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2048,1536,0.00553955551650789
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,2048,0.007274666594134436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2048,5120,0.02188355558448368
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2048,1536,0.013462222284740873
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2048,1024,0.004548444516128964
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,1536,0.14181599352094862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,1536,0.006924444602595435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2048,4096,0.018534221582942538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2048,768,0.0038977778620190094
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2048,1024,0.01310400002532535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,1024,0.006568888823191325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,1024,0.13990133338504368
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2048,512,0.003693333516518275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2048,768,0.0127368892232577
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2048,3584,0.017163554827372234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,768,0.006288888967699475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,768,0.1481928825378418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2048,256,0.0031893334041039148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2048,512,0.012378666963842181
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,512,0.005952888892756567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2048,2560,0.006865777903132969
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,512,0.14176621701982287
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2048,128,0.0031475555151700974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2048,256,0.012106666962305704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,256,0.0058871110280354815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2048,64,0.0031475555151700974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,256,0.14793244997660318
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,2048,32,0.0032097777972618737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,2560,0.007694222033023834
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,2048,128,0.1338720056745741
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1536,65536,0.050359110037485756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2048,64,0.011699555648697747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2048,32,0.011552000211344825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1536,16384,0.01925155520439148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1536,65536,0.1622755527496338
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1536,16384,0.04426133301522997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,16384,0.14110044638315836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1536,12288,0.016224000189039443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,16384,0.0199937770764033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,65536,0.32547465960184735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1536,10240,0.01512977812025282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1536,12288,0.035641776190863714
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,12288,0.14784622192382812
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,12288,0.015487111277050443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1536,8192,0.014183110660976835
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1536,10240,0.03208266695340475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,10240,0.14104355706108943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1536,7168,0.012599111431174807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1536,8192,0.027426666683620874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,8192,0.14652000533209905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1536,7168,0.025175111161337957
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1536,6144,0.011398221883508893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,7168,0.011405333048767514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,7168,0.14164177576700845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,2048,128,0.011697777443461947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,2048,128,0.005938666562239329
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1536,5120,0.010144000252087912
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1536,6144,0.02307022280163235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,6144,0.010657777388890585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,6144,0.14690755473242864
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1536,4096,0.008626666333940294
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1536,5120,0.02035466664367252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,5120,0.009871111147933537
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,5120,0.14442666371663412
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1536,3584,0.007976888782448238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,65536,0.06545599963929918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1536,4096,0.017865777015686035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,4096,0.14290844069586858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,4096,0.008667555948098501
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1536,3072,0.007284444239404466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1536,3584,0.01683555543422699
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,3584,0.008412444757090675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,3584,0.1414791080686781
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,3072,0.00796888851457172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1536,3072,0.016122667325867545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,3072,0.14502755800882974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1536,2048,0.005865777946180767
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1536,2560,0.015150222513410779
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,2560,0.14032355944315592
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,2560,0.007621333003044128
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1536,1536,0.005197333378924264
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1536,2048,0.014096889230940076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,10240,0.014465777410401238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,2048,0.007291555404663086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,2048,0.14327643977271184
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1536,1024,0.00453599997692638
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1536,1536,0.01348533315791024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,1536,0.14187110794915095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,1536,0.0069066666894488865
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1536,1024,0.012813333008024426
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,1024,0.006619555668698416
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,1024,0.1424328883488973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1536,512,0.0037519998020595978
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1536,768,0.012503110700183444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,768,0.006259555617968242
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,768,0.1456800036960178
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1536,512,0.01239466667175293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,512,0.14884621567196316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,512,0.005922666854328579
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1536,256,0.012054222325483957
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1536,128,0.00319377767542998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,256,0.0058808889653947614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1536,128,0.011885333392355176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,128,0.1327928834491306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1536,64,0.0031386667655573953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1536,2560,0.006608888920810487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,128,0.00589155571328269
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1536,64,0.011741333537631564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1536,32,0.011727111207114326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1024,65536,0.037237332926856145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,1536,8192,0.012056888805495368
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1536,768,0.004206222378545337
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1024,65536,0.16383822758992514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1024,16384,0.01574133336544037
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,65536,0.3237840069664849
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,65536,0.06367822488149007
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1024,12288,0.013565333353148567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,16384,0.14684444003634983
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,16384,0.017843556072976854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1024,12288,0.035183111826578774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,12288,0.14518755012088352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,12288,0.014800889624489678
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1024,10240,0.012717333104875354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,10240,0.013426666458447775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1024,10240,0.03159822358025445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,10240,0.14123733838399252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1024,8192,0.013787555197874704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1536,32,0.0031600000543726813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1024,7168,0.012415111064910889
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1024,8192,0.026706667410002813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,8192,0.14021778106689453
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1024,7168,0.024450666374630396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1024,6144,0.011160888605647616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,7168,0.011261333194043903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,7168,0.14578755696614584
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,6144,0.010601777997281818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,6144,0.14114667309655085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1024,5120,0.009738666315873465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1536,256,0.0035031110876136353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1024,5120,0.02029333346419864
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,5120,0.14743555916680232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1024,4096,0.008323555191357931
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,5120,0.009745777481132084
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,1536,256,0.14213421609666613
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1024,3584,0.007294221884674496
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1024,4096,0.01757600075668759
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,4096,0.13915910985734728
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,4096,0.008966222405433655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1024,3072,0.0068604445291890045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,3584,0.00834222220712238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1024,3584,0.01681777834892273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1024,2560,0.006203555398517185
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1024,3072,0.015838222371207345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,3072,0.008032889001899296
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,3072,0.14860888322194418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,8192,0.011760888828171624
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1024,2048,0.0052382221652401825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1024,2560,0.015096889601813423
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,2560,0.007598222129874759
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,2560,0.14226755830976698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1024,1536,0.004729777574539185
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1024,2048,0.014116444521480136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,2048,0.0073626670572492815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,2048,0.14783199628194174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1024,1024,0.0041777777175108595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1024,1536,0.013400889105266996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,1536,0.007045333584149678
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,1536,0.13995999760097927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1024,6144,0.022239999638663396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1024,768,0.003842666745185852
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1024,1024,0.012747555143303342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,1024,0.006562666760550604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1024,512,0.0035413333939181436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,768,0.006246222390068903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1024,768,0.012724444270133972
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1024,16384,0.04417511158519321
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1024,256,0.0031600000543726813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1024,512,0.012087999946541257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,512,0.005943111247486538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,512,0.14673954910702175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1024,128,0.0028320000403457214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1024,256,0.0120328888297081
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,256,0.005888000130653381
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,256,0.13995644781324598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1024,64,0.0028595555987623003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1024,128,0.011393778026103973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,1024,32,0.0028684443483750024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,1024,128,0.005894222193294101
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,128,0.13262222872840032
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,768,65536,0.03191377719243368
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1024,64,0.01180355581972334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,768,16384,0.012771555946932899
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,3584,0.14696178171369764
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,65536,0.06200622187720405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,768,65536,0.16290311018625894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,768,16384,0.04387822084956699
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,16384,0.1465324428346422
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,768,12288,0.012415111064910889
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,16384,0.017872888180944655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,65536,0.3255982134077284
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,768,10240,0.012311110893885294
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,768,12288,0.03562577896647983
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,12288,0.14631556140051946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,12288,0.014825777875052558
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,768,8192,0.010713777608341642
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,768,10240,0.031086222993002996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,10240,0.14035022258758545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,1024,0.14408532778422037
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,768,8192,0.02642400066057841
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,768,7168,0.010072000324726105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,8192,0.011731555892361535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,8192,0.14244533909691706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,1024,768,0.14498755666944715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,768,6144,0.009533333281675974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,768,7168,0.02375288969940609
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,7168,0.011066666907734342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,7168,0.14332444137997097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,768,6144,0.021713778376579285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,6144,0.14286577701568604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,6144,0.010651555326249866
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,768,4096,0.007911110917727152
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,768,5120,0.020265777905782063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,5120,0.1453582180870904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,5120,0.009956444303194681
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,768,3584,0.007248889240953658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,1024,32,0.011385777758227454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,768,4096,0.017504000001483493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,4096,0.14379644393920898
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,768,3072,0.006607999818192587
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,768,3584,0.016701334052615695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,3584,0.008381333616044786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,3584,0.14815733167860243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,768,2560,0.005923555543025334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,768,3072,0.015839111473825242
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,3072,0.007993777592976889
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,3072,0.14238222440083823
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,768,2048,0.005252444495757421
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,768,2560,0.014817777607176038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,2560,0.007679111427730984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,2560,0.1422951089011298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,10240,0.01350044459104538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,768,2048,0.014099554883109199
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,768,1536,0.004846222284767362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,2048,0.007288888924651676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,768,1024,0.0041582224269707995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,768,1536,0.013453333742088742
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,1536,0.00692622239391009
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,1536,0.14437599976857504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,768,768,0.003814222084151374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,768,1024,0.012758221891191272
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,768,5120,0.0090506664580769
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,1024,0.145233776834276
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,768,512,0.003531555748648114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,768,0.006249777972698212
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,768,0.14295732975006104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,768,512,0.012132444315486483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,768,256,0.0031902222997612427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,512,0.006022222340106964
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,512,0.13983733124203152
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,256,0.0059004442559348206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,768,256,0.011731555892361535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,256,0.14210489061143664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,768,128,0.0028533333291610083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,768,128,0.011752000285519494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,128,0.005943999936183293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,768,64,0.0029164445069101122
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,768,32,0.002846222163902389
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,768,64,0.011697777443461947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,2048,0.15007466740078396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,512,65536,0.02567111121283637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,768,32,0.011724444727102915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,512,16384,0.011808888779746162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,1024,0.006594666590293248
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,65536,0.06201066573460897
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,512,65536,0.16197421815660265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,512,12288,0.013370667066839008
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,768,768,0.012716444830099741
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,512,16384,0.04251111216015286
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,65536,0.319032006793552
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,16384,0.017886221408843994
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,512,10240,0.012278222375445895
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,16384,0.14309333430396184
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,512,12288,0.034260445170932345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,12288,0.1445511049694485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,512,8192,0.010687111152542962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,10240,0.013427555561065674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,768,4096,0.00868888861603207
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,10240,0.14084800084431967
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,512,7168,0.009679999616410997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,512,8192,0.02575911084810893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,8192,0.011779555843936073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,8192,0.14798222647772893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,512,6144,0.009335111412737105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,512,7168,0.023697778582572937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,7168,0.14468267228868273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,7168,0.011077333655622272
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,512,5120,0.008396444221337637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,6144,0.010500444306267632
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,768,128,0.13274489508734807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,512,5120,0.020167110694779288
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,5120,0.14352177249060735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,512,4096,0.007658667034573025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,512,4096,0.017490666773584154
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,4096,0.14694844351874456
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,512,3584,0.007245333658324347
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,4096,0.008748444418112436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,512,3584,0.016773333152135212
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,3584,0.14477599991692439
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,12288,0.014933332800865173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,512,3072,0.006574222197135289
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,3584,0.008512000242869059
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,512,10240,0.029826667573716905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,512,2560,0.005897777775923411
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,512,3072,0.015846222639083862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,3072,0.007991111112965478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,3072,0.14464888307783338
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,512,2048,0.005543111099137201
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,512,2560,0.01514755520555708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,2560,0.007601777712504069
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,2560,0.14338666862911648
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,512,1536,0.004860444615284602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,512,6144,0.021872889664438035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,512,2048,0.013813333378897773
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,6144,0.14140710565778944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,2048,0.007251555720965068
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,512,1024,0.004072000169091755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,2048,0.14687821600172254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,1536,0.0069297779765393995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,5120,0.009863110880057016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,512,768,0.0038328886859946777
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,1024,0.006610666712125142
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,512,1024,0.01278133359220293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,1024,0.1420311133066813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,512,512,0.0036524443162812125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,512,768,0.01238044434123569
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,768,0.006263111200597551
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,768,0.1442355579800076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,512,256,0.0032053333189752367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,512,512,0.012060444388124677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,512,0.14421599441104466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,512,128,0.0028826666788922418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,512,256,0.011766222616036734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,256,0.005902222047249476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,256,0.1426195568508572
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,512,64,0.0028435554769304064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,128,0.005892444401979446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,512,32,0.002834666727317704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,512,64,0.011747555600272285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,256,65536,0.01787644459141625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,512,32,0.011559999651379056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,256,65536,0.16127999623616537
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,65536,0.3261431058247884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,512,1536,0.013452444639470843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,65536,0.05482755435837639
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,256,16384,0.041760888364579946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,16384,0.14495821793874106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,256,12288,0.008983111215962304
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,512,512,0.006223111102978389
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,16384,0.017823111679818895
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,256,12288,0.03362133436732822
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,12288,0.014781332678265043
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,12288,0.14041511217753092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,512,128,0.011761777930789523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,128,0.13605156209733751
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,256,10240,0.02971733278698391
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,10240,0.013447110851605734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,10240,0.14321066273583308
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,256,7168,0.007996444072988298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,256,8192,0.025398222936524287
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,8192,0.011730666789743634
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,8192,0.14656978183322483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,256,6144,0.007641777396202087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,256,7168,0.023335110810067918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,7168,0.01111644423670239
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,256,6144,0.021695110532972548
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,6144,0.14258755577935114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,6144,0.01036977767944336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,256,16384,0.01035288886891471
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,512,1536,0.1463057729932997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,256,4096,0.007626666790909237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,256,5120,0.020042666130595736
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,5120,0.1403119961420695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,256,4096,0.01720355616675483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,256,3584,0.0069271110826068455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,4096,0.008955555657545725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,4096,0.1441484424802992
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,256,3584,0.016530666086408827
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,3584,0.008332444561852349
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,256,10240,0.009123555488056606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,3584,0.13982133070627847
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,256,3072,0.006246222390068903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,256,2560,0.00563466673096021
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,256,8192,0.008114666574531132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,3072,0.007994666695594788
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,3072,0.14001866181691489
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,256,2560,0.01479822231663598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,2560,0.14465778403811985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,2560,0.007635555333561367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,256,2048,0.013771555489963956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,7168,0.14455555544959173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,2048,0.14512977335188124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,256,1536,0.004586666822433472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,2048,0.007289778027269576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,256,5120,0.008006222546100616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,256,1536,0.013456000222100152
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,256,1024,0.0038693332009845306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,1536,0.14497689406077066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,5120,0.009941333697901832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,256,1024,0.012772444221708508
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,1024,0.006527110934257507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,256,768,0.01242844429281023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,768,0.1450079944398668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,256,512,0.003173333282272021
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,768,0.006244444598754247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,256,512,0.012040889097584618
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,256,3072,0.015806222955385845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,512,0.14373777972327337
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,256,256,0.0028782222006056043
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,512,0.005950222412745158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,256,2048,0.00519466648499171
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,256,128,0.0028159999185138275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,256,256,0.011740444435013665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,256,0.005909333212508096
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,256,0.14307467142740884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,256,128,0.011689778003427716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,128,0.005903999838564131
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,128,0.13638399706946477
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,256,32,0.0028106667515304354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,128,65536,0.015830222103330824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,256,64,0.011716444459226398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,256,1536,0.0069475554757648045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,256,32,0.011704000333944956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,128,16384,0.009342222577995723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,256,768,0.003555555724435382
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,256,1024,0.14318044980367026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,128,65536,0.1575804418987698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,65536,0.05295644534958733
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,128,12288,0.008329778081840938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,128,16384,0.04082755578888787
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,65536,0.32031822204589844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,16384,0.017883555756674874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,16384,0.14083645078870985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,128,12288,0.03288888931274414
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,12288,0.014818666709793938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,12288,0.145449784066942
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,128,8192,0.008259555531872643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,128,10240,0.029542220963372126
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,10240,0.013375110924243927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,10240,0.14137421713935003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,128,7168,0.00797333319981893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,128,8192,0.02534577747186025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,8192,0.011740444435013665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,8192,0.145468446943495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,128,6144,0.007286222444640265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,128,7168,0.023360000716315374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,7168,0.011072888970375061
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,7168,0.14869156148698595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,128,5120,0.007999111380841997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,128,6144,0.02166044380929735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,6144,0.010393777655230628
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,6144,0.1440675523546007
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,256,64,0.002806222273243798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,128,5120,0.019706666469573975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,5120,0.14744355943467882
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,5120,0.009660444325870937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,128,3584,0.006923555499977536
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,128,4096,0.017455110947291057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,4096,0.008651555412345463
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,4096,0.14250755310058594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,128,3584,0.01682044400109185
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,3584,0.0083155557513237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,3584,0.14309867223103842
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,128,2560,0.006589333216349284
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,3072,0.008094222181373173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,3072,0.146598219871521
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,128,2560,0.014765333798196582
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,128,10240,0.008976000050703684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,2560,0.007637333538797166
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,2560,0.1440862284766303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,128,1536,0.0041991112132867174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,128,2048,0.013764444324705334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,2048,0.007264888948864407
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,2048,0.14035644796159533
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,128,1024,0.0038364442686239877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,128,1536,0.013441777891582914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,1536,0.1452560027440389
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,128,1024,0.01275199982855055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,128,768,0.003535110917356279
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,1024,0.14486666520436606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,1024,0.006559111177921295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,128,4096,0.007266666326257918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,128,768,0.01242844429281023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,768,0.006210666563775804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,128,512,0.0032186667538351486
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,768,0.14475556214650473
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,128,256,0.0028568889117903183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,128,3072,0.006635555376609166
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,512,0.1463706625832452
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,128,256,0.012050666742854647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,128,3072,0.015822221835454304
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,256,0.1486311091317071
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,128,128,0.0028124445428450904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,256,0.005933333188295364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,128,2048,0.006212444355090459
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,128,128,0.01105955574247572
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,128,0.005959110955397288
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,128,64,0.0028071111689011254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,128,32,0.002528888897763358
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,64,65536,0.014504000544548035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,128,64,0.011057777537239922
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,64,16384,0.009696889254781935
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,1536,0.006892444358931647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,64,65536,0.15850755903455946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,64,16384,0.040631999572118126
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,64,12288,0.008323555191357931
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,64,10240,0.007681777907742395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,64,8192,0.0069573331210348345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,64,12288,0.03288088904486762
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,64,10240,0.02933688958485921
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,64,7168,0.006616000086069107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,64,6144,0.006650666809744305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,64,8192,0.025220443805058796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,64,5120,0.006285333385070165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,64,7168,0.02331200076474084
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,128,512,0.012063110868136087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,64,6144,0.02162222233083513
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,64,4096,0.006626666833957036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,64,5120,0.01962488889694214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,64,3072,0.006272888845867581
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,64,4096,0.01722755531469981
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,64,3584,0.016647110382715862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,64,2560,0.006592888798978593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,64,2048,0.006239999913507038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,64,2560,0.01481244464715322
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,64,1536,0.004242666479614046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,64,2048,0.013798221945762634
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,64,1024,0.0038168889780839286
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,128,32,0.011404444773991903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,64,1536,0.013440888788965015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,64,768,0.003547555456558863
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,64,1024,0.012709333664841123
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,64,768,0.012482666307025485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,64,256,0.0028497777465316984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,64,512,0.012100444071822695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,64,128,0.002583111118939188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,64,256,0.01180622229973475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,64,64,0.002567111204067866
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,64,128,0.011721777419249216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,64,32,0.002495999965402815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,256,128,512,0.005964444329341252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,64,64,0.011752888560295105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,32,65536,0.014095111025704278
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,64,3584,0.006800888727108638
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,32,16384,0.008976889153321585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,32,12288,0.008347555167145198
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,32,16384,0.04016177852948507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,64,3072,0.015497777197096082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,32,65536,0.15729422039455837
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,32,10240,0.007967111137178209
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,32,12288,0.0328053335348765
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,32,8192,0.006986666884687211
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,32,7168,0.006622222148709827
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,32,6144,0.006588444527652528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,32,10240,0.029304888513353135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,32,8192,0.025024889243973628
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,32,7168,0.02331377731429206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,32,4096,0.005976000179847081
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,32,3584,0.006640000061856375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,32,6144,0.02160888910293579
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,32,5120,0.0195777780479855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,32,3072,0.006265777680608962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,64,512,0.0031439999325407874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,32,4096,0.017234666479958426
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,32,2560,0.006072888771692912
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,32,2048,0.006239999913507038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,32,3584,0.01667555504375034
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,32,1536,0.004226666771703296
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,32,3072,0.015480000111791821
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,32,1024,0.0038364442686239877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,32,2560,0.014787554740905762
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,32,768,0.003555555724435382
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,32,2048,0.013755555782053204
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,32,512,0.0032053333189752367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,32,1536,0.013437333206335703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,32,256,0.002798222212327851
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,32,128,0.0028248888750871024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,32,1024,0.012072888513406118
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,32,64,0.0025048889219760895
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,32,512,0.011751111182901593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,32,256,0.012096000214417776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,32,32,0.0024951110697454875
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,32,128,0.011704000333944956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,32,64,0.0110871113008923
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,32,32,0.011749332977665795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,65536,16384,0.3908702267540826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,256,32,5120,0.006129777679840724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,16384,0.14054489135742188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,65536,12288,0.29719021585252553
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,65536,12288,0.2234711117214627
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,32,768,0.012392889294359418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,256,64,32,0.011423110961914062
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,12288,0.10769333442052205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,65536,10240,0.23690488603379992
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,65536,16384,0.29777778519524467
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,65536,10240,0.1883617771996392
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,16384,0.34574490123324925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,65536,8192,0.198707554075453
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,65536,8192,0.15568089485168457
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,12288,0.266536898083157
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,65536,7168,0.17071199417114258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,8192,0.07339644432067871
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,65536,7168,0.13630755742390951
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,10240,0.229423099093967
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,65536,6144,0.1467155615488688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,7168,0.06531200144026014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,8192,0.19364799393547905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,6144,0.05703733364741007
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,65536,5120,0.13116799460517034
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,10240,0.08973955445819431
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,7168,0.17563288741641572
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,65536,5120,0.09901244772805108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,6144,0.15655822224087185
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,5120,0.04827733172310723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,65536,4096,0.08229066928227742
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,65536,3584,0.08886577685674031
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,5120,0.1504862176047431
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,4096,0.04041688972049289
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,4096,0.14555644989013672
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,3584,0.03581510980923971
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,65536,3584,0.07345777750015259
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,65536,3072,0.08053688870535956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,65536,6144,0.11886310577392578
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,65536,3072,0.06557777855131361
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,65536,2560,0.06428444385528564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,3072,0.0321342216597663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,3072,0.14775555663638645
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,65536,2560,0.057877335283491343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,65536,2048,0.05394488904211256
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,65536,4096,0.10319822364383274
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,2560,0.028203556934992473
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,65536,2048,0.05017066664165921
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,2560,0.14559377564324275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,2048,0.02436622149414486
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,65536,1024,0.030926220946841772
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,1536,0.01909244391653273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,1536,0.14617510636647543
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,65536,1024,0.034280889564090304
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,65536,768,0.02385866641998291
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,1024,0.015847111741701763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,65536,768,0.02916533417171902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,768,0.14663822121090359
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,3584,0.14840977721744114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,65536,512,0.01724177764521705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,65536,512,0.024617777930365667
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,65536,1536,0.04099999864896139
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,512,0.01332800007528729
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,65536,256,0.01235466698805491
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,65536,1536,0.041703111595577665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,65536,256,0.022291556000709534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,65536,128,0.010227555202113258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,256,0.1418399943245782
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,256,0.01183111137813992
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,65536,128,0.02163022259871165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,128,0.13223200374179417
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,1024,0.14656355645921496
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,65536,64,0.008918222453859117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,128,0.011726222104496427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,65536,32,0.0099653336736891
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,65536,64,0.021399110555648804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,65536,768,0.014138667119873894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,16384,65536,0.38091201252407497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,512,0.1453404426574707
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,16384,16384,0.1039537787437439
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,16384,65536,0.3844346735212538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,16384,16384,0.10366044441858928
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,65536,2048,0.14251466592152914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,16384,0.041907555527157254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,16384,12288,0.07647199763192071
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,16384,12288,0.08006044228871663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,12288,0.03331911232736375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,16384,10240,0.06620177957746717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,12288,0.14735821882883707
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,65536,32,0.02498488956027561
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,16384,10240,0.06826400094562106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,16384,8192,0.05291111270586649
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,10240,0.02886755598915948
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,16384,8192,0.05924533473120796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,16384,7168,0.04664355516433716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,8192,0.024085332949956257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,8192,0.1479520003000895
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,16384,7168,0.052654223309622876
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,65536,0.14961777793036565
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,65536,0.45374398761325413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,16384,6144,0.04470933477083842
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,16384,0.1503902276357015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,6144,0.14102755652533636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,6144,0.019612444771660697
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,16384,5120,0.03550488750139872
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,16384,5120,0.03955644369125366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,10240,0.14102578163146973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,5120,0.14066133234235975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,5120,0.017487999465730455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,16384,4096,0.029162665208180744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,16384,4096,0.03423110975159539
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,16384,6144,0.04085244403945075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,7168,0.14496533075968424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,4096,0.14032444689008924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,16384,3584,0.02613777750068241
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,4096,0.015255999233987598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,16384,3072,0.023445333043734234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,16384,3584,0.0314044455687205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,3584,0.014019555515713163
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,3584,0.14207200209299722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,16384,2560,0.01965777741538154
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,16384,3072,0.028769777880774602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,3072,0.1466008954577976
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,16384,2560,0.025157334076033697
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,2560,0.1400231122970581
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,2560,0.01204711116022534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,16384,2048,0.02260355485810174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,2048,0.14494666788313124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,2048,0.011109333071443768
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,16384,1536,0.020224000016848247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,7168,0.02225333286656274
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,16384,1024,0.010663111176755695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,1536,0.14159022437201604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,1536,0.009003555609120263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,16384,1024,0.018144889010323417
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,1024,0.008033778104517195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,1024,0.1409697797563341
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,16384,768,0.008965333302815756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,3072,0.013088000317414602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,16384,768,0.016467556357383728
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,16384,512,0.007268444531493717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,768,0.0075395554304122925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,768,0.14070933394961888
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,16384,2048,0.01645599967903561
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,16384,512,0.014856000741322836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,512,0.14813600646124944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,16384,256,0.014134221606784396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,16384,128,0.004256000121434529
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,256,0.006908444480763541
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,16384,1536,0.013064000341627332
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,16384,128,0.014088888963063559
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,16384,64,0.004538666870858935
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,128,0.13808266321818033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,128,0.006642666541867786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,16384,32,0.004369777937730153
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,16384,64,0.013794666363133324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,16384,32,0.014330666926172046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,12288,65536,0.2773759894900852
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,12288,65536,0.29669067594740123
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,12288,16384,0.07345955239401923
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,65536,0.13349421819051108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,12288,16384,0.08220711019304064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,16384,512,0.007039111521508958
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,12288,12288,0.05694933070076836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,16384,0.03979111048910353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,16384,0.1481297810872396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,16384,256,0.14931821823120117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,12288,12288,0.0640888876385159
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,12288,10240,0.04895022180345324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,12288,10240,0.05503466725349426
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,12288,0.14726132816738552
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,12288,8192,0.0399528874291314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,10240,0.026897778113683064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,10240,0.14463910791609022
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,12288,8192,0.04832977718777127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,12288,7168,0.035411556561787925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,8192,0.022312000393867493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,8192,0.14746044741736516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,65536,0.44323023160298664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,12288,6144,0.030777778890397813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,7168,0.020432000358899433
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,12288,7168,0.04320799973275927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,7168,0.143251551522149
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,12288,5120,0.026749332745869953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,16384,256,0.005294222384691238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,12288,6144,0.0365608897474077
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,12288,5120,0.03235466612709893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,5120,0.14644000265333387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,12288,4096,0.022643556197484333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,5120,0.016866667403115165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,12288,4096,0.027934223413467407
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,4096,0.0148053334818946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,4096,0.14546755949656168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,12288,3584,0.02565244502491421
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,12288,0.031129777431488037
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,12288,3072,0.017511111166742113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,3584,0.01346577786737018
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,3584,0.14021778106689453
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,12288,2560,0.014841778410805596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,12288,3072,0.023652444283167522
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,3072,0.012716444830099741
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,3072,0.14299644364251032
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,12288,2048,0.01277955538696713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,12288,2560,0.02159022291501363
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,2560,0.14880888991885716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,12288,2048,0.020222221811612446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,2048,0.14041244983673096
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,2048,0.009664889011118147
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,6144,0.018571555614471436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,6144,0.1462284459008111
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,12288,1536,0.018207111292415194
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,12288,1024,0.008959111240175035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,1536,0.14505244625939265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,1536,0.00866844422287411
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,12288,3584,0.01997155613369412
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,12288,768,0.00757155567407608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,12288,1024,0.01606666710641649
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,1024,0.007930667036109501
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,1024,0.14521333906385633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,12288,512,0.006227555374304454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,12288,768,0.015174221661355762
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,768,0.1428791152106391
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,768,0.007583111524581909
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,12288,256,0.004556444370084339
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,12288,512,0.014148443937301636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,512,0.0069368887278768755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,512,0.1400231122970581
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,12288,128,0.003999111139112049
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,12288,256,0.013153778182135688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,256,0.006596444381607904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,2560,0.011413333316644033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,256,0.14552888605329725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,12288,64,0.003540444291300244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,12288,32,0.0036657779581016968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,12288,128,0.006629333313968446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,12288,64,0.013481777575280933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,12288,1536,0.010764444039927589
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,12288,32,0.013757333159446716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,10240,65536,0.26354755295647514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,10240,16384,0.07039466831419203
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,10240,65536,0.27771910031636554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,65536,0.12969688574473062
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,10240,16384,0.07977688974804349
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,16384,0.03825511203871833
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,16384,0.1477715571721395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,10240,12288,0.05453955464892917
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,10240,12288,0.061899556054009326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,12288,0.0298977792263031
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,12288,128,0.013227555486891003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,10240,10240,0.04551733202404446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,12288,128,0.14035111003451878
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,12288,0.14586844709184435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,10240,8192,0.038400888442993164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,10240,10240,0.0532915559079912
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,10240,0.025621333056026038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,10240,8192,0.046987556748920016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,10240,0.14452621671888563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,65536,0.4416488806406657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,10240,7168,0.033824000093672014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,256,128,128,0.1393102275000678
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,8192,0.02165422174665663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,8192,0.14838043848673502
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,10240,5120,0.025723555021815833
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,10240,7168,0.041476445065604314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,7168,0.14188622103797063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,6144,0.01795111099878947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,6144,0.14270933469136557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,10240,4096,0.021753778060277302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,10240,5120,0.03118488854832119
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,5120,0.1444044510523478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,10240,4096,0.026763555076387193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,10240,3584,0.01902577777703603
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,4096,0.014327110515700446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,4096,0.1441564427481757
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,10240,3072,0.016743110285864938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,10240,3584,0.024957334001859028
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,3584,0.013191999660597907
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,10240,3072,0.0229759994480345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,3072,0.012558221817016602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,3072,0.1420231130388048
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,10240,2560,0.021458667185571458
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,2560,0.14229244656032985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,10240,2048,0.012721777790122561
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,2560,0.010752889017264048
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,10240,6144,0.029767112599478826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,10240,2048,0.01959733333852556
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,10240,1536,0.010665777656767102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,10240,6144,0.035078220897250705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,2048,0.14800000190734863
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,2048,0.00926755534278022
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,10240,1024,0.008712000317043727
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,10240,1536,0.017838221457269456
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,5120,0.01624000072479248
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,1536,0.14206133948432073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,10240,1024,0.015837333268589444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,10240,768,0.006950221955776215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,1024,0.007626666790909237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,1024,0.1419502231809828
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,10240,512,0.005603555589914322
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,3584,0.14520800113677979
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,768,0.1425564421547784
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,10240,512,0.014025777578353882
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,10240,2560,0.014584889014561972
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,512,0.14620888233184814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,10240,256,0.013183111117945777
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,256,0.1449644433127509
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,256,0.006643555644485686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,10240,128,0.0041999999019834726
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,7168,0.019654222660594516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,10240,128,0.013096888860066732
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,1536,0.008339555727110969
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,10240,128,0.13317867120107016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,128,0.006602666858169768
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,10240,64,0.0038542221817705366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,10240,32,0.0038657776183552216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,10240,64,0.01311288856797748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,10240,32,0.013439999686347114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,10240,768,0.014473777678277759
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,768,0.007272889216740926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,8192,16384,0.051065779394573636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,10240,256,0.0047644442982143825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,10240,512,0.0069777775141927935
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,8192,65536,0.24017333984375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,65536,0.08157955275641547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,8192,16384,0.06738577948676215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,8192,12288,0.04114222195413377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,16384,0.025511110822359722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,8192,10240,0.03366933266321818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,16384,0.1410613324907091
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,8192,12288,0.05374044511053297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,12288,0.020244444410006206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,12288,0.14354933632744685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,8192,10240,0.04723822077115377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,10240,0.017695999807781644
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,10240,0.14055466651916504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,8192,8192,0.04070399867163764
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,8192,0.015495111544926962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,8192,0.14335732989841038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,65536,0.30529599719577366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,8192,7168,0.03582311007711623
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,7168,0.01441422270403968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,8192,65536,0.19351022773318818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,8192,6144,0.030664887693193223
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,8192,5120,0.01937599976857503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,6144,0.13980621761745876
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,8192,5120,0.02700888944996728
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,8192,4096,0.01718577742576599
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,5120,0.012348444097571902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,4096,0.14012800322638616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,8192,4096,0.02397511071628994
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,8192,3584,0.014303111367755465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,8192,8192,0.029179556502236262
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,4096,0.010985777609878115
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,8192,3584,0.02268444498380025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,3584,0.13912444644504124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,3584,0.010198222266303169
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,8192,3072,0.020974222156736586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,8192,6144,0.022468444373872545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,3072,0.1411235597398546
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,8192,2560,0.011240888800885944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,3072,0.009348444640636444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,8192,2560,0.019536000159051683
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,2560,0.1475777758492364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,8192,2048,0.0099653336736891
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,2560,0.008325333396593729
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,8192,2048,0.018059555027219985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,2048,0.14116177293989393
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,5120,0.14582044548458523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,8192,1536,0.008636444807052612
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,2048,0.007697777615653143
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,8192,1024,0.006907555378145642
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,8192,1536,0.016529776983790927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,1536,0.14426578415764704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,8192,1024,0.014466666513019137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,8192,768,0.005703111075692707
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,1024,0.14111910925971136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,8192,768,0.013768889009952545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,8192,7168,0.025404444999165002
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,8192,512,0.004530666602982415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,768,0.006562666760550604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,8192,512,0.013454222016864352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,7168,0.14578666951921251
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,8192,256,0.0038382220599386427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,512,0.14049243927001953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,512,0.006272000157170826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,6144,0.013163555827405719
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,8192,128,0.0035075553589397003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,256,0.14019644260406494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,256,0.005936000082227919
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,128,0.005945777727497949
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,8192,128,0.012440889245933957
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,128,0.13153778182135686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,8192,64,0.0038364442686239877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,8192,32,0.0038319999972979226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,1536,0.007170666423108842
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,8192,64,0.012714666624863943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,8192,32,0.012820444173283048
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,7168,65536,0.17816532982720268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,8192,1024,0.006723555425802867
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,7168,16384,0.04831288920508491
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,8192,3072,0.013064000341627332
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,8192,768,0.14173955387539333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,7168,16384,0.06755733489990234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,65536,0.07837422026528253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,7168,12288,0.03787822193569607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,16384,0.024910221497217815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,16384,0.13927822642856175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,7168,12288,0.05293066634072197
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,7168,10240,0.032087110810809664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,12288,0.14013866583506265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,12288,0.019426667028003268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,7168,8192,0.02754666739039951
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,7168,10240,0.04608177807595995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,8192,256,0.013064000341627332
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,10240,0.017151999804708693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,10240,0.1385591162575616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,7168,7168,0.02423288921515147
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,65536,0.29975732167561847
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,7168,8192,0.03901155458556281
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,8192,0.01516355574131012
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,7168,6144,0.020729778541458976
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,8192,0.1454275581571791
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,7168,7168,0.03477955526775784
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,7168,0.013827555709415011
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,7168,0.140538665983412
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,7168,5120,0.01792888840039571
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,7168,6144,0.029951112137900457
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,6144,0.012920888761679331
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,7168,4096,0.015477332803938123
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,6144,0.14061778121524388
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,5120,0.01198755535814497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,5120,0.14012089040544298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,7168,5120,0.026610665851169165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,7168,3584,0.013647110925780402
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,4096,0.010327999790509542
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,7168,4096,0.02360800074206458
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,7168,65536,0.23448001013861763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,7168,3584,0.02244088881545597
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,3584,0.14472088548872206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,3584,0.009369778136412302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,7168,2560,0.010979555547237396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,7168,3072,0.02095111045572493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,3072,0.008716444174448649
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,7168,2048,0.009341333475377824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,7168,2560,0.01953155630164676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,2560,0.007983999947706858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,2560,0.14144978258344862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,7168,1536,0.00793688909875022
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,7168,2048,0.017839999662505258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,2048,0.007430222299363878
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,7168,1536,0.015981333123313058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,7168,1024,0.006555555595291986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,7168,1024,0.014448000325096978
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,1024,0.14324177636040583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,4096,0.14546399646335179
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,1024,0.006608888920810487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,7168,3072,0.012318222059143914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,7168,768,0.005555555638339784
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,7168,768,0.013743110828929476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,768,0.006291555447710886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,7168,512,0.004519111166397731
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,3072,0.148634672164917
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,7168,512,0.013452444639470843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,7168,256,0.0038648889296584656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,512,0.006280888699822956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,2048,0.13876356018914118
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,7168,256,0.01276177747382058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,1536,0.007028444773621029
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,256,0.005951111101441913
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,7168,128,0.0124177775449223
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,128,0.1323946714401245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,7168,64,0.00350577798154619
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,7168,128,0.005902222047249476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,7168,32,0.0035093331502543557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,7168,64,0.012508444488048553
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,7168,32,0.013053333593739403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,6144,65536,0.16627555423312718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,6144,65536,0.1964071061876085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,768,0.14318845007154676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,6144,16384,0.0450799994998508
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,65536,0.07618221971723768
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,512,0.14433511098225912
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,6144,16384,0.05695911248524984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,16384,0.023591111103693645
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,16384,0.1413360039393107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,1536,0.1402133305867513
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,6144,12288,0.04571911030345493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,7168,128,0.003546666767862108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,6144,10240,0.030093332131703694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,65536,0.28636709849039715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,12288,0.14532800515492758
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,12288,0.018907555275493197
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,6144,8192,0.025433777107132807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,6144,10240,0.0391244457827674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,10240,0.14528711636861166
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,10240,0.01668444441424476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,6144,7168,0.02126400007141961
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,6144,8192,0.03345955411593119
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,8192,0.14019644260406494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,6144,6144,0.019511111908488803
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,8192,0.014695111248228284
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,6144,7168,0.030173334810468886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,7168,0.013460444079505073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,6144,5120,0.01669066647688548
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,6144,6144,0.02626755502488878
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,6144,0.01278222186697854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,6144,0.14576621850331625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,6144,4096,0.014648000399271647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,6144,5120,0.02369155486424764
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,5120,0.011432888607184092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,5120,0.14609511693318686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,6144,3584,0.012687111066447364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,6144,4096,0.021635555558734473
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,4096,0.00982844415638182
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,6144,3584,0.020435555113686454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,6144,3072,0.011668444507651858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,6144,12288,0.03367822368939718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,7168,256,0.14587644735972086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,3072,0.1448106633292304
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,6144,2560,0.010022222167915767
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,3072,0.008793777889675563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,6144,2560,0.01777688827779558
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,2560,0.14886844158172607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,6144,2048,0.009376889301670922
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,2560,0.007947555846638149
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,6144,2048,0.016516443755891588
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,2048,0.14534489313761392
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,2048,0.007532444265153673
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,4096,0.14123111300998265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,6144,1536,0.014874666929244995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,3584,0.009294222626421187
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,1536,0.007054222126801808
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,1536,0.14195288552178278
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,6144,3072,0.01919911139541202
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,6144,1024,0.013805333111021252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,1024,0.14180800649854872
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,6144,768,0.005576000031497743
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,1024,0.006627555522653792
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,6144,768,0.013413333230548434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,768,0.006247111078765657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,768,0.14699910746680364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,6144,512,0.004200889004601372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,7168,0.14062400658925375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,512,0.006269333263238271
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,6144,512,0.013087111214796701
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,6144,256,0.0038488888078265717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,6144,256,0.012424888710180918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,256,0.005939555664857228
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,6144,128,0.0035111109415690103
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,256,0.14604000250498453
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,6144,1536,0.007692444655630324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,6144,64,0.0031999999450312722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,6144,128,0.00590133335855272
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,128,0.13458043999142116
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,6144,32,0.0032008888406885993
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,6144,64,0.012079111403889127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,3584,0.1441404422124227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,6144,32,0.01239999963177575
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,5120,65536,0.1456106636259291
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,5120,16384,0.04689688815010918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,5120,65536,0.19315111637115479
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,65536,0.07356977462768555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,5120,16384,0.056440002388424344
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,5120,12288,0.03602666656176249
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,16384,0.022405332989162866
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,16384,0.14173244105445013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,5120,10240,0.030265778303146362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,5120,12288,0.04365333252482944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,12288,0.018019555343521964
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,6144,512,0.14729866716596815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,65536,0.2844177881876628
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,5120,8192,0.025740444660186768
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,5120,10240,0.03751733236842685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,10240,0.016142222616407607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,10240,0.14537155628204346
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,5120,7168,0.022703111171722412
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,8192,0.014130666851997375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,8192,0.1469751066631741
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,5120,8192,0.03236622280544705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,6144,128,0.012049777640236748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,5120,6144,0.01995733380317688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,5120,7168,0.029390222496456567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,5120,5120,0.016905777984195285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,7168,0.145505772696601
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,5120,6144,0.025722665919197932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,6144,0.012079999678664737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,6144,1024,0.006205333189831839
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,5120,4096,0.01457244488928053
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,5120,5120,0.023418666587935552
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,5120,0.010771555205186209
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,5120,0.14296089278327093
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,5120,3584,0.01274577776590983
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,5120,4096,0.021243555678261652
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,4096,0.009492444495360056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,4096,0.14682311481899685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,5120,3072,0.011504888534545898
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,5120,3584,0.019819556011093985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,3584,0.008736000292830998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,3584,0.14373332924313015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,5120,3072,0.01882488860024346
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,5120,2560,0.01017599966790941
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,3072,0.1463822258843316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,5120,2048,0.00868266655339135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,5120,2560,0.017588444881969027
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,3072,0.008271111382378472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,2560,0.14721866448720297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,5120,2048,0.016176000237464905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,5120,1536,0.007823111282454597
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,2048,0.00755555596616533
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,5120,1536,0.014800889624489678
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,7168,0.013162666724787818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,1536,0.14646400345696345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,5120,1024,0.005558222118351195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,1536,0.006962666908899943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,6144,0.1412924395667182
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,5120,768,0.005211555709441503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,5120,1024,0.013635555903116861
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,1024,0.006625777731339137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,1024,0.1395297712749905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,5120,512,0.004192000048028098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,5120,768,0.013395555317401886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,768,0.006236444330877728
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,768,0.14767555395762125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,5120,256,0.0038204445607132385
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,5120,512,0.012776888906955719
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,512,0.006105777704053455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,5120,256,0.012085333466529846
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,256,0.13905866940816244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,5120,128,0.003540444291300244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,256,0.005936000082227919
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,12288,0.14204266336229113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,2560,0.00792266676823298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,5120,64,0.0031724443866146933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,5120,128,0.012328888807031842
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,128,0.13910488287607828
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,5120,64,0.012373333175977072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,2048,0.1440755526224772
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,5120,32,0.012481778032249875
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,4096,65536,0.10937155617607965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,4096,16384,0.03537155522240533
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,4096,65536,0.1852337784237332
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,65536,0.07119199964735243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,4096,16384,0.05361510978804695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,16384,0.02122933334774441
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,4096,12288,0.030569778548346624
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,16384,0.141393780708313
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,65536,0.28031733300950795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,4096,12288,0.04149333304829068
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,4096,10240,0.02536533276240031
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,12288,0.01759822169939677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,12288,0.14811555544535318
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,5120,512,0.1408311128616333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,4096,8192,0.022297778063350256
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,10240,0.015821332732836407
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,4096,10240,0.03603377938270569
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,10240,0.14544444613986546
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,4096,8192,0.031502223677105375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,4096,7168,0.018541332748201158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,8192,0.1473591062757704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,8192,0.013766222529941134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,4096,6144,0.01751733322938283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,5120,32,0.0035146665241983202
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,4096,7168,0.02903555499182807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,7168,0.012536889149083031
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,7168,0.1417591174443563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,4096,5120,0.014422222971916199
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,4096,6144,0.02538577715555827
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,6144,0.1475564506318834
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,4096,4096,0.012700444294346703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,4096,5120,0.022953778505325317
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,5120,0.010036444498433007
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,5120,0.14098666773902044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,4096,3584,0.009322666459613377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,4096,4096,0.02083555526203579
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,4096,0.009015999734401703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,4096,0.14265510771009657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,4096,3584,0.019653333557976615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,3584,0.144923554526435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,4096,2560,0.007968000239796108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,4096,3072,0.017768000562985737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,3072,0.008285333712895712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,3072,0.14105244477589926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,4096,2048,0.0068417779273457
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,4096,2560,0.016525333126386006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,2560,0.007912889122962952
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,2560,0.14236444897121853
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,4096,1536,0.005991111199061076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,4096,2048,0.015471110741297403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,2048,0.14546399646335179
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,4096,1536,0.014588443769348992
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,1536,0.14053600364261204
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,4096,1024,0.004920000003443824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,5120,128,0.005883555445406172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,4096,1024,0.013463111387358772
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,1024,0.14175111717647976
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,4096,768,0.00423555572827657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,6144,0.011370666325092316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,4096,768,0.013083555632167392
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,768,0.006314666734801398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,4096,512,0.0038720000949170855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,768,0.14005155033535427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,4096,3072,0.008839111361238692
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,4096,512,0.012753778033786349
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,512,0.005950222412745158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,3584,0.008640000389681922
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,512,0.1438159942626953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,256,0.005881777654091517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,256,0.1401617791917589
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,4096,128,0.0031724443866146933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,4096,128,0.012068444656001197
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,4096,128,0.13438933425479466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,2048,0.007309333317809635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,4096,64,0.0034862222770849862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,128,0.005883555445406172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,1536,0.0069644442862934535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,4096,64,0.01202488856183158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,4096,32,0.012395555774370829
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3584,65536,0.12810577286614314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,4096,1024,0.006619555668698416
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3584,16384,0.03933511177698771
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3584,65536,0.18248177899254692
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,65536,0.07012622223960029
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3584,16384,0.05286488930384318
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,16384,0.1463386615117391
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3584,12288,0.02798488736152649
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,16384,0.02108088963561588
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,65536,0.28091822730170357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,4096,256,0.01239111108912362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,12288,0.017253332667880587
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3584,10240,0.024998222788174946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,12288,0.14599822627173528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3584,8192,0.020208888583713107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3584,10240,0.03561066587766012
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,10240,0.01553066737122006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,10240,0.1456408898035685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3584,8192,0.030857778257793848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3584,7168,0.018147556318177115
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,4096,32,0.0032088889016045462
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,8192,0.14279288715786403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,8192,0.013412444127930535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3584,6144,0.017041777571042378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3584,7168,0.028460443019866943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,7168,0.14419289429982504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,7168,0.012415111064910889
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3584,5120,0.01443288889196184
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3584,6144,0.025106666816605464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,6144,0.011032889286677042
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,6144,0.14236888620588514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3584,4096,0.012014221813943652
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3584,5120,0.02276533345381419
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,5120,0.010265777508417765
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,5120,0.14429956012301975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3584,3584,0.011659555964999728
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3584,4096,0.020249777370029025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,4096,0.009128000173303816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,4096,0.1443262232674493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3584,3584,0.019306666321224637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,4096,256,0.0035217776894569397
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,3584,0.008676444490750631
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3584,2560,0.007802666889296637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3584,3072,0.017508443858888414
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,3072,0.14379555649227566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,3072,0.008284444610277811
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3584,2048,0.006608888920810487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3584,12288,0.04055466585689121
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3584,2560,0.016181333197487723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,2560,0.1449751059214274
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3584,1536,0.005619555711746216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3584,2048,0.015120888749758402
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,2048,0.00739733295308219
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,2048,0.14452266693115234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3584,1024,0.004560889055331548
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3584,1536,0.014271999398867289
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,1536,0.0069439998931354946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,1536,0.14590933587816027
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3584,768,0.0041511112617121804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3584,1024,0.01344622257683012
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,1024,0.14172177844577366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3584,512,0.0038782221575578055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3584,768,0.01312622262371911
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,768,0.006224888894293044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,768,0.14110399617089167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3584,256,0.0035377778112888336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3584,512,0.012746666868527731
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,512,0.005942222144868638
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,512,0.1463057729932997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3584,128,0.003173333282272021
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3584,256,0.012411555482281579
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,256,0.005889777921968036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3584,3072,0.008363555702898238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3584,128,0.012070222033394707
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3584,64,0.003152888889114062
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,3584,0.1399386723836263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,128,0.13211911254458958
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,128,0.0058515556156635284
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3584,64,0.012113778127564324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3584,32,0.012064889073371887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3072,65536,0.0935795572068956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3072,16384,0.030108445220523413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,2560,0.007880888879299164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3072,65536,0.1450586716334025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,65536,0.06914577881495158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3072,12288,0.021252445048756067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3072,16384,0.04291466540760464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,16384,0.02072000006834666
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,16384,0.14543999565972224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,3584,1024,0.006613333192136552
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3072,12288,0.03293066554599338
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3072,10240,0.018553778529167175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,12288,0.1402719948026869
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3072,10240,0.02930755416552226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,10240,0.0151573336786694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,10240,0.14653155538770887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3072,8192,0.015487111277050443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,3584,256,0.14062400658925375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3072,8192,0.02613777750068241
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,8192,0.1456800036960178
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3072,7168,0.013960000541475085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3584,32,0.003140444349911478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3072,7168,0.02461155586772495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,7168,0.011782222323947482
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,7168,0.1467591126759847
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3072,6144,0.02157955533928341
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,6144,0.010715555813577441
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,6144,0.14232888486650255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3072,5120,0.010987555815113915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3072,5120,0.01975022256374359
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,5120,0.009973333113723332
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,5120,0.14107555813259548
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3072,4096,0.017825777331988018
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,12288,0.01724444495307075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,4096,0.14053689108954534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,65536,0.28005777464972603
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3072,3584,0.008977777428097194
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,4096,0.008955555657545725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3072,3072,0.008281778130266402
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,3584,0.008654221892356873
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,3584,0.1454391082127889
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3072,3072,0.015505777464972602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,3072,0.00796088907453749
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3072,2560,0.007622222105662028
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3072,6144,0.012359110845459832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,2560,0.14057332939571804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3072,2048,0.006737777756320105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,2560,0.007644444704055786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3072,2048,0.014132445057233175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3072,1536,0.005528000079923206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,2048,0.14309688409169516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,2048,0.007277333074145847
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3072,1536,0.013734222286277346
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,1536,0.1423457728491889
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3072,1024,0.013068444199032254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3072,3584,0.01685066686736213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,1024,0.18158133824666342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3072,768,0.0042035554846127825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,1024,0.006584888945023219
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,3072,0.14397866196102566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3072,768,0.01276177747382058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3072,512,0.0038124442928367188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3072,2560,0.01512888901763492
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3072,512,0.012429333395428128
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,512,0.1434897714191013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3072,256,0.0032035555276605818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,512,0.00620977787507905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,1536,0.0069022224181228215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3072,1024,0.004226666771703296
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3072,128,0.003170666595300039
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,256,0.005888888819350137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3072,128,0.011774222056070963
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,128,0.13777066601647273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3072,64,0.003154666680428717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,128,0.0058906666106647914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3072,32,0.0028533333291610083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3072,64,0.011697777443461947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3072,32,0.011743110915025076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2560,65536,0.07395378086302015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,768,0.00629066675901413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2560,65536,0.14089066452450222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,65536,0.06809510787328084
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2560,16384,0.029672887590196397
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2560,16384,0.04153066542413499
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,3072,256,0.012083555261294046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,65536,0.2773706648084852
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,16384,0.1411786609225803
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2560,12288,0.024102222588327196
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,256,0.14181689421335855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,16384,0.020596444606781006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2560,10240,0.016883555385801528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,12288,0.13979732990264893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,12288,0.01683911184469859
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2560,8192,0.014223999447292753
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2560,10240,0.028800888193978205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,10240,0.1426711082458496
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,10240,0.014793778459231058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2560,7168,0.013167111410035027
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2560,8192,0.025736888249715168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,8192,0.01236088905069563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,8192,0.14449332820044622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,3072,8192,0.013047999805874295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2560,6144,0.011797332929240333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2560,7168,0.024174221687846716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,7168,0.011391110718250275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,7168,0.145487109820048
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,3072,768,0.14151199658711752
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2560,5120,0.010693333215183683
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2560,6144,0.020954666866196528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,6144,0.14520977603064641
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,6144,0.010600888894663917
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2560,4096,0.009101333717505137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2560,5120,0.01947111056910621
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,5120,0.14155111047956678
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,5120,0.010004444254769219
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2560,3584,0.008716444174448649
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,4096,0.008949333594905006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2560,3584,0.01612977849112617
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2560,3072,0.007928888830873702
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,3072,4096,0.009999111294746399
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2560,3072,0.01514755520555708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,3072,0.14662755860222712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2560,12288,0.03165600034925673
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,3072,0.008086221913496653
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2560,2560,0.014434667097197639
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,2560,0.14036177264319524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2560,2048,0.006229333165619109
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,2560,0.007644444704055786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2560,2048,0.01401511165830824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,2048,0.1443493366241455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2560,1536,0.005201777650250329
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,2048,0.007303111255168915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2560,4096,0.01717155509524875
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2560,1536,0.013406222065289816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,4096,0.14611111746893987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2560,1024,0.004303110970391167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,1536,0.14132977856530085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,1536,0.006956444846259222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2560,768,0.0041831110914548235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,3584,0.008339555727110969
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2560,1024,0.012816000315878125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,1024,0.006581333362393909
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2560,512,0.0038480001191298165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2560,768,0.01276088919904497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,768,0.006279110908508301
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,768,0.14505688349405924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2560,2560,0.007268444531493717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2560,256,0.003527111063400904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2560,512,0.01237155579858356
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,512,0.006253333141406377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,512,0.13989333311716715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2560,128,0.003173333282272021
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2560,256,0.011778666741318174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,256,0.005899555567238066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,256,0.14477511246999106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2560,64,0.003160888950030009
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2560,128,0.011791999969217511
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2560,32,0.0031999999450312722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,2560,128,0.00591822216908137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,128,0.13603111108144125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2048,65536,0.05984444750679863
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2560,32,0.011711111499203576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2048,16384,0.022895110978020564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2048,65536,0.1371199952231513
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,65536,0.06666488779915704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2048,16384,0.04001244571473863
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2048,12288,0.019561777512232464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,16384,0.02035199933581882
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,3584,0.14123733838399252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2048,12288,0.03180711136923896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,2560,1024,0.1451466613345676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,12288,0.14096889230940077
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2048,10240,0.01751377847459581
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,12288,0.016157334049542744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2048,8192,0.014136888914638095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,10240,0.013815110756291283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2048,10240,0.028314666615592107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,10240,0.13986044459872776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2048,7168,0.012700444294346703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2048,8192,0.02540622154871623
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,8192,0.012079111403889127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,8192,0.13994756009843615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2048,7168,0.023312888211674158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,7168,0.14065066973368326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,7168,0.01144444445768992
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2048,5120,0.010317333042621613
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2048,6144,0.020999999509917364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2560,64,0.011744000017642975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,6144,0.14315199851989746
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,6144,0.010671999719407825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2048,4096,0.008999110923873054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2048,5120,0.018597332967652213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,5120,0.009824000298976898
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2048,3584,0.008264888491895463
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2048,4096,0.017097777790493436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,4096,0.14358933766682944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,4096,0.008972444468074376
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2048,3072,0.007918222082985772
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2048,3584,0.016042666302786935
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,16384,0.1485111051135593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,3584,0.008582221964995066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,3584,0.14032799667782253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2048,2560,0.006605333338181178
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2048,3072,0.015132443772421943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,3072,0.008266666697131263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,3072,0.1465217802259657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2048,2560,0.014635556273990206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2048,2048,0.005582222094138463
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,2560,0.14356356196933323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,2560,0.007655999726719326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2048,1536,0.004860444615284602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2048,2048,0.013800889253616333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,2048,0.14079022407531738
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,2048,0.007311999797821045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2048,1024,0.004190222256713443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2048,1536,0.01308444473478529
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,1536,0.0069520001610120135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2048,6144,0.01162577751610014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2048,768,0.0041617775956789655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2048,1024,0.012871999707486896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,1024,0.006572444405820634
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2048,768,0.012373333175977072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2048,512,0.0038311113086011675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,768,0.14389067225986057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,5120,0.14034933514065212
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2048,512,0.0120319997270902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,512,0.14269777139027914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,512,0.0059262220230367445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,65536,0.2774311171637641
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2048,256,0.01200799975130293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2048,128,0.003154666680428717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,256,0.13982844352722168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,256,0.005931555396980709
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2048,128,0.011739555332395764
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,128,0.13060711489783391
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2048,64,0.003142222141226133
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2048,32,0.0028684443483750024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2048,64,0.011738667057620155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,2048,32,0.01202488856183158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,1536,0.14695378144582114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1536,65536,0.05224711034033033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,2048,1024,0.14172089099884033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1536,16384,0.018044445249769423
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,768,0.006275555739800136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,65536,0.27711378203497994
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1536,16384,0.038564443588256836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1536,12288,0.016454221473799813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,16384,0.01943733294804891
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,2048,256,0.00351200004418691
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1536,10240,0.015062222878138224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1536,12288,0.031204445494545832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,12288,0.015557333827018738
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,12288,0.14501155747307673
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1536,8192,0.013788444300492605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1536,10240,0.02779200010829502
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,10240,0.013438222308953604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,10240,0.14261333147684732
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1536,7168,0.013061333033773633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1536,8192,0.024957334001859028
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,2048,128,0.005922666854328579
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,8192,0.14432799816131592
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,8192,0.01180266671710544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1536,6144,0.011727111207114326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1536,7168,0.022405332989162866
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,7168,0.011081777513027191
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,7168,0.14578933186001247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1536,5120,0.010041777458455827
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,6144,0.010610666539933948
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,6144,0.14524355199601915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1536,4096,0.00905600024594201
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1536,5120,0.018415111634466384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,5120,0.14311022228664821
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,5120,0.00999644481473499
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1536,3584,0.00834222220712238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1536,65536,0.13532622655232748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1536,4096,0.016536888149049547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,4096,0.009002666506502364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,4096,0.1448177761501736
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1536,3072,0.007274666594134436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,3584,0.008446222378147973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1536,3584,0.01587199999226464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,16384,0.147507561577691
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1536,2560,0.006588444527652528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1536,3072,0.015046222342385186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,3072,0.14387910895877415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,3072,0.007992888490358988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1536,2048,0.005571555760171678
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1536,2560,0.014166222678290473
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,2560,0.007651555869314406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1536,1536,0.004860444615284602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1536,2048,0.013751111096805997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,2048,0.007300444775157505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,2048,0.14564177725050184
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1536,1024,0.0042053332759274375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1536,1536,0.013159111142158508
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,1536,0.006966222491529252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,1536,0.1436515516704983
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1536,768,0.0038817777401871155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1536,1024,0.012773333324326409
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,1024,0.006632888896597757
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,1024,0.14556977483961317
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1536,6144,0.019888000355826486
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1536,768,0.0123857781291008
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,768,0.006248000181383557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,768,0.1464115513695611
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1536,256,0.003177777760558658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1536,512,0.012104888757069906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,512,0.14496889379289415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,65536,0.06525599956512451
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1536,256,0.012062222593360476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1536,128,0.0031840000301599503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,256,0.14077777332729763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,256,0.0058764442801475525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,3584,0.14047822687360975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1536,64,0.002855111120475663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,128,0.1316035588582357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,128,0.005917333480384614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1536,32,0.0028124445428450904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,1536,2560,0.141628450817532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1536,64,0.011747555600272285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1024,65536,0.038359999656677246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1536,32,0.011734222372372946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1024,16384,0.01551555593808492
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,65536,0.06318755282296075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1024,65536,0.13170044951968723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1024,16384,0.03829155696762933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1024,12288,0.013568888935777875
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,65536,0.27711465623643666
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,16384,0.13868621985117593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,16384,0.017861333158281114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1536,512,0.003495111233658261
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1024,10240,0.012607110871209038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1024,12288,0.030786666605207656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,12288,0.14291022883521184
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,12288,0.014771555860837301
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1024,8192,0.011120888921949597
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1024,10240,0.027523555689387854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,10240,0.013426666458447775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,10240,0.14101244343651667
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,1536,512,0.005974222388532426
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1024,7168,0.01070755554570092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1024,8192,0.02438666588730282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1024,6144,0.009642666412724389
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,8192,0.1430399947696262
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1024,7168,0.02195466723706987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,7168,0.011126221881972419
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,7168,0.14689155419667563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1024,5120,0.008945778012275696
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1024,6144,0.019323555959595572
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1536,128,0.011716444459226398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,6144,0.010431999961535135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1024,4096,0.008289777570300633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1024,5120,0.01789955629242791
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,5120,0.009700444837411245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1024,3584,0.007302222152551015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1024,4096,0.01648622254530589
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,4096,0.008671110702885522
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,4096,0.1447884506649441
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1024,3072,0.0069724445541699724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1024,3584,0.015809777710172866
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,3584,0.00832266691658232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,3584,0.1417902178234524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1024,3072,0.014813333749771118
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,3072,0.00795111142926746
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,3072,0.14073422220018175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1024,2560,0.01421333352724711
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1024,2048,0.005252444495757421
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,2560,0.007640889121426477
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,2560,0.14765689108106825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1024,2048,0.013784888717863293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,2048,0.007305777735180325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,2048,0.14092089070214167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,8192,0.011735999749766456
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1024,1536,0.013016888664828407
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,1536,0.139284438557095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1024,1024,0.004229333251714706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,6144,0.1411315600077311
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1024,1024,0.012704888979593912
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,5120,0.14081066184573704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1024,768,0.0038248888320393036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,1024,0.00655377780397733
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1024,512,0.003504000190231535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1024,768,0.01242844429281023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,768,0.006287999865081575
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,768,0.13804266187879774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1024,256,0.0032035555276605818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1024,512,0.012096000214417776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1024,2560,0.005949333310127258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,512,0.14064178201887342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1024,128,0.0028693332440323303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1024,256,0.011687111523416309
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,256,0.00591111100382275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,256,0.14467022154066297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1024,64,0.002874666617976295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1024,128,0.011744000017642975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,128,0.00591822216908137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1024,32,0.0028293333533737394
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,128,0.13293245103624132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,768,65536,0.03180799881617228
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1024,64,0.011696889168686338
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,1024,1536,0.004912888838185204
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,1024,32,0.011698666546079846
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,768,16384,0.01314222233162986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,768,65536,0.13047999805874294
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,65536,0.0616168909602695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,768,16384,0.037767112255096436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,1536,0.0069120000633928515
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,16384,0.1448168887032403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,65536,0.2773919900258382
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,16384,0.017889777819315594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,768,10240,0.013287111288971372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,768,12288,0.030481778913074072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,1024,1024,0.14888621701134577
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,12288,0.01478488908873664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,768,8192,0.01202933324707879
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,768,10240,0.026720888084835474
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,10240,0.013444444371594323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,10240,0.14228088325924343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,768,7168,0.011352000137170156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,768,8192,0.023084445132149592
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,8192,0.011724444727102915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,1024,512,0.005948444621430502
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,8192,0.14022933112250432
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,768,6144,0.009993777506881291
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,768,7168,0.021721776988771226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,7168,0.011050666371981302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,7168,0.1407466729482015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,768,5120,0.009331555830107795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,6144,0.1421688927544488
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,768,5120,0.017893332574102614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,5120,0.14584622118208143
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,768,4096,0.00832444429397583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,5120,0.009735110733244155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,768,4096,0.016342222690582275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,4096,0.00884622252649731
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,4096,0.14084445105658636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,768,3584,0.007988444632954067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,768,12288,0.011390222443474663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,768,3072,0.006351999938488007
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,768,3584,0.015836444165971544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,3584,0.008338666624493068
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,3584,0.13945244418250188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,768,2560,0.0059164443777667145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,12288,0.14469067255655924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,768,3072,0.0148053334818946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,3072,0.007985778152942657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,3072,0.14187644587622747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,768,2048,0.005398222141795688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,768,2560,0.014182221558358936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,2560,0.007656888829337225
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,2560,0.1457173294491238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,768,1536,0.004781333522664176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,768,2048,0.013801777528391944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,2048,0.14768089188469782
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,768,1024,0.004200889004601372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,768,1536,0.01313066648112403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,1536,0.0069164443347189165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,1536,0.14345244566599527
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,768,768,0.003822222352027893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,768,6144,0.019539554913838703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,768,1024,0.012426666915416718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,1024,0.006607999818192587
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,1024,0.14481155077616373
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,768,512,0.0033111110743549135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,768,768,0.012278222375445895
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,768,0.00628355559375551
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,768,0.14021244313981798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,768,256,0.0031893334041039148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,768,512,0.012376888758606382
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,512,0.005932444499598608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,512,0.1417822175555759
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,768,256,0.011712888876597086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,256,0.14818488226996526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,256,0.005899555567238066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,768,64,0.0028640000770489373
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,768,128,0.011747555600272285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,128,0.005934222290913264
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,768,32,0.0028275555620590844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,768,128,0.130840884314643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,512,65536,0.02419111132621765
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,768,64,0.011696000066068439
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,768,32,0.01111644423670239
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,512,16384,0.013904000322024027
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,512,65536,0.1279253297381931
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,65536,0.06052711274888781
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,65536,0.27432534429762095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,512,16384,0.03695377707481384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,2048,0.007258666886223688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,16384,0.017865777015686035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,16384,0.14209155241648355
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,512,12288,0.012072888513406118
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,512,10240,0.01071466671095954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,512,12288,0.029531555043326482
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,12288,0.014807111687130399
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,512,10240,0.026360889275868733
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,10240,0.013458666702111563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,10240,0.14274577299753824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,768,6144,0.01035199976629681
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,512,7168,0.008850666383902231
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,512,8192,0.022908444205919903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,8192,0.011759999725553723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,8192,0.13968177636464438
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,512,7168,0.02161688937081231
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,512,6144,0.008281778130266402
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,7168,0.1421991056866116
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,7168,0.011158222125636207
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,6144,0.01037333326207267
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,6144,0.14177066749996609
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,512,5120,0.017856889300876193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,5120,0.14671644899580213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,512,4096,0.007679999702506595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,5120,0.009852444132169088
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,512,4096,0.016524444023768108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,4096,0.14429688453674316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,512,3584,0.007303111255168915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,4096,0.00867733359336853
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,512,3584,0.015816888875431485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,12288,0.14138933022816977
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,512,8192,0.009689778089523315
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,512,3072,0.006552888701359431
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,3584,0.00850933293501536
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,512,3072,0.01479288935661316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,3072,0.13997244834899902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,512,2560,0.005619555711746216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,512,5120,0.007995555798212687
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,512,6144,0.019312888383865356
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,512,2560,0.0144177774588267
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,512,2048,0.005207111024194294
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,2560,0.14871822463141546
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,2560,0.007718222008811102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,512,1536,0.004555555681387584
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,2048,0.007302222152551015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,2048,0.14075911045074463
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,512,1536,0.013119999733236102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,512,1024,0.0038844446341196695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,1536,0.1478382216559516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,1536,0.0069546666410234236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,512,1024,0.0127697777416971
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,512,768,0.003548444559176763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,3584,0.1460097763273451
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,512,768,0.01236977759334776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,768,0.00628266649113761
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,512,512,0.003491555651028951
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,768,0.14553688632117376
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,3072,0.008028444316652086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,768,128,0.0028648889727062653
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,512,512,0.012084444363911947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,512,0.006257777826653586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,512,0.14638133843739828
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,512,256,0.011771555576059552
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,256,0.005936000082227919
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,512,128,0.0027875554644399216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,512,2048,0.013501333693663279
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,512,128,0.011742221812407175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,128,0.13437778419918484
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,512,64,0.011743110915025076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,512,32,0.0027831111931138565
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,512,32,0.011744000017642975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,256,65536,0.021225778592957392
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,256,65536,0.12793510489993626
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,65536,0.2752995491027832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,256,16384,0.010017777482668558
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,65536,0.053564445839987866
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,256,16384,0.03581510980923971
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,256,0.1447546614540948
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,256,12288,0.00865955568022198
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,16384,0.017861333158281114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,128,0.005903111149867375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,512,64,0.0028000000036425064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,256,12288,0.028914666838116113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,256,10240,0.008991999758614434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,12288,0.01479822231663598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,12288,0.1421066655053033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,512,1024,0.006568000134494569
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,256,8192,0.008676444490750631
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,512,1024,0.14699110719892713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,256,10240,0.02572800053490533
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,10240,0.013398222625255585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,256,7168,0.007975999679830339
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,256,8192,0.02273599969016181
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,8192,0.011739555332395764
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,8192,0.14303110705481634
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,256,6144,0.007627555893527136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,256,7168,0.021307556165589228
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,7168,0.14489689138200548
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,7168,0.011063999599880643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,256,6144,0.01921777758333418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,6144,0.010382222632567087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,6144,0.1457751062181261
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,256,4096,0.007314666277832455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,256,5120,0.017687999539905123
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,5120,0.009697777529557547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,512,256,0.003133333391613431
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,256,3584,0.007256000406212277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,256,4096,0.016193777322769165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,4096,0.00866844422287411
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,256,3584,0.015449777245521545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,16384,0.1425902181201511
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,3584,0.00832266691658232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,256,3072,0.006773333168692059
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,3072,0.007999999655617608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,256,3072,0.01478844384352366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,3072,0.14128177695804173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,256,2560,0.005589333259397083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,256,2560,0.014127999544143677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,10240,0.14524355199601915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,2560,0.007611555357774098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,256,2048,0.00518844442235099
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,2560,0.13972711563110352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,256,2048,0.013774221969975365
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,2048,0.007266666326257918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,2048,0.14456266827053493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,256,1536,0.004298666699065102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,256,5120,0.008025777836640676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,256,1024,0.003922666526503033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,256,1536,0.013067555924256643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,1536,0.14410489135318333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,256,1024,0.01274577776590983
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,5120,0.14069778389400905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,256,768,0.003519111209445529
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,1024,0.006719111154476802
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,1024,0.13965333832634821
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,256,512,0.003145777723855443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,4096,0.14138221740722656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,256,768,0.012120889292822944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,768,0.1450559960471259
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,256,512,0.012053333222866058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,256,256,0.003168888803985384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,3584,0.14744355943467882
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,512,0.005953777581453323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,512,0.14024799399905735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,256,128,0.0025884444928831523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,256,256,0.011940444509188334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,256,0.14388088385264078
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,256,64,0.002830222249031067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,256,128,0.13430755668216282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,128,0.005899555567238066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,256,32,0.002492444382773505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,128,65536,0.01516711049609714
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,256,64,0.011717333561844297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,256,32,0.011693333586057028
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,128,65536,0.1245759990480211
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,65536,0.054639998409483165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,128,16384,0.009639999932712978
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,65536,0.27361954583062065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,1536,0.0069511110583941145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,128,12288,0.008294222255547842
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,128,16384,0.037187556425730385
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,16384,0.14145067003038195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,16384,0.017811555001470778
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,128,10240,0.009010666774378883
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,128,12288,0.028875556257035997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,12288,0.014787554740905762
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,12288,0.14077244864569768
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,768,0.006233777850866318
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,128,8192,0.00832711077398724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,10240,0.013415999710559845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,128,8192,0.022506665852334764
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,8192,0.1421413289176093
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,256,256,0.005937777873542573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,8192,0.011759111450778114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,256,128,0.011416888899273343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,128,6144,0.007592889169851939
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,128,7168,0.021309332715140447
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,7168,0.011043555206722684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,7168,0.14763200283050537
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,128,5120,0.0070062221752272705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,128,6144,0.019140443868107267
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,6144,0.14099466800689697
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,128,5120,0.01731911136044396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,128,4096,0.0069724445541699724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,5120,0.14013866583506265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,128,4096,0.01587199999226464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,4096,0.14043466250101724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,128,3584,0.00630044440428416
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,4096,0.008676444490750631
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,128,3584,0.015797333584891427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,3584,0.00830666638082928
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,128,3072,0.006541333264774746
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,10240,0.14465067121717665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,128,7168,0.007993777592976889
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,3072,0.008000888758235509
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,128,2560,0.006575111299753189
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,128,2560,0.01428711083200243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,2560,0.14520800113677979
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,128,2048,0.006197333335876465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,6144,0.01034755590889189
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,2560,0.007612444460391998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,128,2048,0.013423110875818463
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,2048,0.007302222152551015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,5120,0.009697777529557547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,2048,0.14323911401960585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,128,1536,0.01311555587583118
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,1536,0.006940444310506185
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,128,1024,0.003843555433882607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,128,1024,0.012431999875439538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,1024,0.14476799964904785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,128,768,0.0035084444615576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,3584,0.14174400435553655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,1024,0.006573333508438534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,128,768,0.012395555774370829
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,768,0.006298666612969504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,128,3072,0.014470222923490735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,128,512,0.0031893334041039148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,768,0.14050132698482937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,3072,0.14189778433905706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,128,256,0.002879111096262932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,512,0.005943999936183293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,128,512,0.012090666426552666
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,512,0.14887911743587917
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,128,256,0.011788444386588203
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,128,128,0.002850666642189026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,256,0.14045333862304688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,256,0.005952000204059813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,128,64,0.0025351111673646504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,128,128,0.011706666813956367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,192,128,128,0.005937777873542573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,128,32,0.0024906665914588505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,64,65536,0.01423911088042789
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,128,1536,0.004523555437723796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,128,32,0.011693333586057028
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,1536,0.14151822196112737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,64,65536,0.12576533688439265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,64,12288,0.007967111137178209
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,64,16384,0.03495288888613383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,64,10240,0.007294221884674496
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,64,8192,0.006949333681000604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,64,12288,0.028798222541809082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,64,10240,0.025779555241266888
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,64,7168,0.00676711110605134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,64,8192,0.022631110416518316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,64,6144,0.006566222343179915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,64,5120,0.006205333189831839
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,64,7168,0.02124533388349745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,64,4096,0.005919999960396025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,64,6144,0.01923288901646932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,64,3584,0.006239111224810283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,64,4096,0.016188444362746347
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,64,3072,0.006274666637182236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,64,3584,0.015743111570676167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,64,2560,0.006558222075303395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,64,3072,0.014535999960369535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,64,2048,0.005917333480384614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,64,2560,0.014149333039919535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,128,64,0.011715555356608497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,64,1536,0.004503111044565837
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,64,1024,0.003822222352027893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,64,2048,0.01344622257683012
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,64,1536,0.012835555606418185
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,64,768,0.0035377778112888336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,64,512,0.0031519999934567344
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,64,1024,0.012412444584899478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,64,256,0.0028008888992998335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,64,768,0.012092444631788464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,64,128,0.0025253333151340485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,64,512,0.012038222617573209
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,128,10240,0.02573066618707445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,64,64,0.0028071111689011254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,64,256,0.012071111136012606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,64,32,0.002529777793420686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,32,65536,0.0147724449634552
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,64,128,0.011371555427710215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,64,64,0.011695110963450538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,64,32,0.011393778026103973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,32,12288,0.007972444097201029
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,64,5120,0.017536888519922893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,32,65536,0.12337777349683975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,32,16384,0.034994665119383074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,32,10240,0.007602666815121968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,32,12288,0.02886311213175456
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,32,7168,0.006893333461549547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,32,10240,0.02571022179391649
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,32,6144,0.00656444455186526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,32,8192,0.022685334086418152
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,32,7168,0.021288888321982488
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,32,5120,0.00628266649113761
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,32,4096,0.005904888941182031
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,32,6144,0.01903199983967675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,64,16384,0.008984888593355814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,32,3584,0.006545777950021956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,32,5120,0.017834666702482436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,32,3072,0.005974222388532426
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,32,4096,0.016178665889634024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,32,2560,0.006284444282452266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,32,2048,0.005952000204059813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,32,3584,0.015451555450757345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,32,3072,0.014702222413486905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,32,1536,0.004540444248252445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,32,2560,0.014140443669425117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,32,2048,0.013404444687896304
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,32,768,0.0035546666218174826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,32,1536,0.013047999805874295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,32,1024,0.012717333104875354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,32,768,0.012068444656001197
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,32,256,0.002806222273243798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,32,128,0.002855111120475663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,32,512,0.012038222617573209
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,32,16384,0.008617777791288164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,32,64,0.0025066667132907445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,32,32,0.002508444504605399
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,32,256,0.011687999798191918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,32,128,0.011690666278203329
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,32,64,0.011426666544543372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,192,32,32,0.011401777466138205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,32,8192,0.0069395556218094295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,65536,16384,0.34892890188429093
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,65536,16384,0.29319821463690865
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,65536,12288,0.26653512318929035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,16384,0.1379786729812622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,32,1024,0.0036755556033717263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,65536,12288,0.21949244870079887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,192,32,512,0.0032142222755485107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,65536,10240,0.2322915660010444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,65536,10240,0.1856400039460924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,12288,0.10348355770111084
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,65536,8192,0.19593332873450386
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,65536,8192,0.152911106745402
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,10240,0.08919288714726765
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,65536,7168,0.1674248907301161
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,12288,0.2615751160515679
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,8192,0.070997330877516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,16384,0.3409173223707411
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,65536,7168,0.13221600320604113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,65536,6144,0.14857867028978136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,7168,0.06315288941065471
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,65536,6144,0.11430044968922932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,8192,0.19134222136603463
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,65536,5120,0.12160355514950222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,65536,5120,0.0977413323190477
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,7168,0.17187733120388457
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,6144,0.15644799338446722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,65536,4096,0.1015555527475145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,5120,0.0473599996831682
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,65536,3584,0.08735911051432292
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,65536,4096,0.08156710863113403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,4096,0.039691554175482854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,5120,0.14657333162095812
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,65536,3072,0.07687644163767497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,10240,0.22293334537082246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,4096,0.14300978183746338
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,3584,0.14451200432247585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,3072,0.0317884451813168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,65536,2560,0.06283022297753228
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,3072,0.1441404422124227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,65536,2560,0.05704533391528659
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,65536,2048,0.05230311221546597
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,2560,0.02765422231621212
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,2560,0.14408089054955378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,65536,2048,0.049508445792728
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,65536,1536,0.04018755422698127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,2048,0.02401688860522376
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,2048,0.14522666401333278
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,65536,1536,0.040660444233152605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,65536,3584,0.07221777571572198
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,3584,0.03529511226548089
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,1536,0.018920888503392536
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,1536,0.14404088921017116
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,65536,768,0.02276977731121911
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,65536,1024,0.033398220936457314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,1024,0.015491555134455362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,65536,3072,0.06390488810009427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,1024,0.1435342232386271
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,65536,512,0.016128889388508268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,65536,768,0.029556443293889362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,6144,0.055333332882987134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,768,0.0140346669488483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,768,0.14555733733707005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,65536,256,0.010431999961535135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,512,0.14699555767907038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,65536,512,0.025052444802390203
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,65536,128,0.009097778134875828
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,65536,256,0.02254844374126858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,256,0.14620978302425808
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,65536,64,0.008275555239783393
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,65536,128,0.021518222159809534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,65536,128,0.13777600394354925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,65536,32,0.00871288859181934
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,128,0.011516444385051727
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,65536,64,0.02163200080394745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,65536,32,0.021960000197092693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,16384,65536,0.38195021947224933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,65536,1024,0.030108445220523413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,16384,65536,0.3646586736043294
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,16384,16384,0.09669599930445354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,16384,16384,0.10070221953921848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,65536,0.1473919947942098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,16384,12288,0.07570577992333306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,512,0.013263111313184103
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,16384,12288,0.07666489150789049
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,16384,0.14728711711035833
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,65536,256,0.011728000309732227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,16384,10240,0.06331288814544678
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,12288,0.03313422203063965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,16384,10240,0.06703555583953857
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,12288,0.14311733510759142
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,10240,0.02868266569243537
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,10240,0.14088177680969238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,16384,8192,0.05760000149408976
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,8192,0.02422222163942125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,16384,7168,0.046615110503302686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,8192,0.14362043804592556
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,16384,6144,0.04028622309366862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,16384,7168,0.04827022221353319
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,7168,0.022062222162882488
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,7168,0.1459439992904663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,16384,5120,0.03456000155872769
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,16384,6144,0.04351199997795952
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,65536,0.4343760013580322
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,6144,0.019536889261669584
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,6144,0.14545067151387533
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,16384,0.04199733336766561
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,16384,5120,0.038482666015625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,16384,4096,0.029097778929604426
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,5120,0.017255999975734286
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,5120,0.1432808902528551
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,16384,3584,0.025557332568698462
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,16384,4096,0.03346488873163859
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,4096,0.015134221977657743
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,4096,0.14793954955206975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,16384,3072,0.022999111149046157
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,16384,8192,0.05422222283151415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,3584,0.014120888378885059
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,3584,0.14284888903299967
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,16384,2560,0.01924622224436866
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,16384,3072,0.027219555444187585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,3072,0.013089777694808112
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,3072,0.14588088459438747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,16384,2560,0.024706666668256123
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,16384,2048,0.016159999701711867
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,2560,0.14176800515916613
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,16384,1536,0.013044444223244986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,16384,2048,0.02232711182700263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,2048,0.011058666639857821
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,2048,0.1408888896306356
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,16384,1024,0.01037333326207267
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,16384,1536,0.020240000552601285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,1536,0.00906311141120063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,1536,0.1467884381612142
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,16384,1024,0.01776088939772712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,16384,768,0.008947555389669206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,1024,0.1405528916252984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,1024,0.007998222278224098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,16384,512,0.007271111011505127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,16384,768,0.01738133364253574
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,768,0.14431911044650608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,768,0.007954667011896769
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,16384,256,0.004872000051869286
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,16384,512,0.015443555182880826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,512,0.007330666813585494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,512,0.1408124499850803
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,16384,128,0.004219555606444676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,256,0.006911110960774952
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,256,0.1438159942626953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,16384,128,0.014071111877759298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,16384,64,0.003871110992299186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,128,0.0069324444565508105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,16384,128,0.1373653279410468
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,16384,32,0.00388711111413108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,16384,2560,0.012040889097584618
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,16384,64,0.013788444300492605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,16384,32,0.014485332700941296
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,12288,65536,0.2761431005265978
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,12288,16384,0.07209866576724581
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,12288,65536,0.2818257808685303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,12288,16384,0.07903289132648043
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,65536,0.13203733497195774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,16384,0.039247999588648476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,12288,12288,0.056800888644324414
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,16384,0.14580889542897543
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,16384,256,0.014134221606784396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,12288,12288,0.060652441448635526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,16384,3584,0.030937777625189886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,12288,0.0310808883772956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,12288,10240,0.05392533209588793
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,12288,0.145087997118632
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,12288,8192,0.039585779110590615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,10240,0.02664622167746226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,10240,0.14480888843536377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,12288,7168,0.034832000732421875
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,12288,8192,0.04714133342107137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,8192,0.022405332989162866
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,8192,0.14472356107499865
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,12288,7168,0.03939377930429246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,7168,0.020276443825827706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,7168,0.14238844977484807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,12288,6144,0.03559911251068115
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,65536,0.42220266660054523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,12288,5120,0.026354667213228013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,6144,0.018414222531848483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,6144,0.14389867252773708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,12288,4096,0.022893332772784766
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,12288,5120,0.03160711129506429
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,5120,0.016522667474216886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,12288,3584,0.01976799964904785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,12288,4096,0.027863999207814533
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,4096,0.014577777849303352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,4096,0.14120622475941977
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,12288,3072,0.01753866672515869
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,12288,3584,0.02534577747186025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,3584,0.013510222236315409
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,12288,10240,0.04735022121005588
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,3584,0.14444000191158718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,12288,2560,0.01479644411140018
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,12288,3072,0.023023110296991136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,3072,0.14191555976867676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,3072,0.012720888687504662
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,12288,2048,0.012724444270133972
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,12288,2560,0.021320000290870667
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,2560,0.011403555671374003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,2560,0.14344888263278538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,12288,1536,0.010431999961535135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,12288,2048,0.019908444748984445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,2048,0.14275199837154812
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,2048,0.00981955561372969
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,12288,1024,0.008960888617568547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,12288,6144,0.030395554171668157
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,12288,1536,0.018207111292415194
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,1536,0.008357333640257517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,12288,768,0.007607999775144789
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,12288,1024,0.015858666764365304
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,1024,0.007977777885066139
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,1024,0.14336444271935359
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,12288,512,0.005551110953092575
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,12288,768,0.01516177753607432
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,768,0.007409777906205919
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,5120,0.14089600245157877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,512,0.007024889190991719
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,12288,256,0.004558222161398994
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,12288,256,0.01369066701995002
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,256,0.140864888827006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,12288,128,0.0038746665749284956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,12288,128,0.013450666434235044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,128,0.1335173315472073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,1536,0.14672088623046875
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,12288,64,0.00416088890698221
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,128,0.0069217777086628815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,12288,32,0.004156444635656145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,12288,64,0.013094222380055321
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,768,0.1420853270424737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,12288,32,0.01407199932469262
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,12288,512,0.014391111003028022
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,12288,512,0.14177600542704263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,10240,16384,0.06548088788986206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,10240,65536,0.2461573282877604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,10240,65536,0.2678728898366292
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,10240,12288,0.051502221160464816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,65536,0.12946755356258816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,16384,0.037858668300840594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,16384,0.14546755949656168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,10240,12288,0.059303998947143555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,12288,0.0295466681321462
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,10240,10240,0.042697777350743614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,12288,0.14502755800882974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,10240,10240,0.05231555634074741
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,10240,8192,0.036885334385765925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,10240,0.02573866645495097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,10240,0.1414088937971327
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,10240,8192,0.045257776975631714
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,10240,7168,0.03177244464556376
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,8192,0.02164622147878011
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,8192,0.14361866315205893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,10240,6144,0.027820444769329492
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,10240,7168,0.037752889924579196
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,7168,0.019613333874278598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,65536,0.4211511082119412
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,10240,6144,0.033979554971059166
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,6144,0.017848889032999676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,10240,5120,0.024248000648286607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,6144,0.14663110838996038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,10240,4096,0.020145777199003432
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,10240,16384,0.07643377780914307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,5120,0.016169778174824186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,5120,0.14741244581010607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,10240,3584,0.017879999346203275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,4096,0.014356444279352823
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,4096,0.14170222812228733
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,10240,3584,0.02426133387618595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,10240,3072,0.01605244477589925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,3584,0.1410399940278795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,10240,3072,0.022626666559113398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,3072,0.14141421847873265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,10240,2560,0.01365066650840971
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,3072,0.012039999994966718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,10240,2560,0.021006221572558086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,7168,0.1421111159854465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,10240,2048,0.012055111428101858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,2560,0.010701333483060202
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,10240,2048,0.01958311100800832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,12288,256,0.006607111129495833
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,2048,0.1402275562286377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,10240,5120,0.030174222257402208
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,10240,1536,0.01036888857682546
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,2048,0.00923466682434082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,10240,1024,0.008543999658690559
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,10240,1536,0.017838221457269456
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,1536,0.00832444429397583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,1536,0.14552799860636392
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,10240,768,0.006882666713661618
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,10240,1024,0.015490666031837463
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,3584,0.013191111385822296
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,1024,0.14227467113071018
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,10240,768,0.014753777119848462
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,10240,512,0.00553688903649648
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,768,0.14038666089375815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,768,0.00740799970097012
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,10240,512,0.014078223043017916
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,10240,256,0.004519111166397731
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,512,0.0069644442862934535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,512,0.14114133516947427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,10240,256,0.013388444152143268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,256,0.14180889394548205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,256,0.006689777804745569
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,2560,0.14575910568237305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,192,128,128,0.14366666475931802
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,10240,64,0.003866666720973121
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,10240,128,0.012735111018021902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,128,0.006593777901596493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,10240,128,0.13184000386132136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,10240,32,0.0036720000207424164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,10240,64,0.012827555338541666
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,10240,32,0.013080000049538083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,8192,65536,0.17999111281500924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,10240,4096,0.026150221625963848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,8192,16384,0.049903111325369946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,8192,65536,0.22622399859958223
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,10240,1024,0.007636444436179266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,65536,0.08131111330456205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,8192,12288,0.03938222262594435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,16384,0.025399999486075506
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,16384,0.140773336092631
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,8192,12288,0.05163822240299649
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,8192,10240,0.03254755669169956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,12288,0.1458968851301405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,8192,10240,0.045237332582473755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,10240,128,0.0038693332009845306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,8192,8192,0.02794755498568217
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,10240,0.017654221918847825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,10240,0.13987822002834743
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,8192,7168,0.0240666667620341
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,65536,0.2835306591457791
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,8192,8192,0.03910577628347609
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,8192,0.015257777439223396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,8192,0.14536266856723362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,8192,6144,0.021338666478792827
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,8192,7168,0.03302133414480422
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,7168,0.01423377792040507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,7168,0.14589422278934056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,8192,5120,0.01833777791923947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,8192,6144,0.029254221253924902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,6144,0.013163555827405719
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,6144,0.1456177764468723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,8192,4096,0.016318221886952717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,8192,5120,0.026457778281635706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,5120,0.012068444656001197
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,8192,3584,0.013787555197874704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,8192,16384,0.06590488884184095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,4096,0.14109333356221518
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,4096,0.01108000013563368
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,8192,3072,0.012488889197508493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,8192,3584,0.02193244463867611
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,3584,0.010017777482668558
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,3584,0.14456978109147814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,12288,0.02023555503951179
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,8192,2560,0.010889777706729041
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,8192,3072,0.020668443706300523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,3072,0.009286222358544668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,8192,2048,0.009621333744790819
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,8192,2560,0.019375110665957134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,2560,0.008028444316652086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,2560,0.1446968846850925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,8192,1536,0.008040889269775814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,8192,2048,0.0181688881582684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,2048,0.14182666937510172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,8192,1536,0.016014221641752455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,1536,0.14004533820682102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,8192,1024,0.006656000183688269
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,1536,0.007282666862010956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,8192,1024,0.01481066644191742
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,1024,0.006599999964237213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,5120,0.1436764399210612
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,8192,768,0.014112888111008538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,8192,512,0.004335111214054955
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,768,0.14163111315833196
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,768,0.0064106666379504735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,8192,512,0.013435555828942193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,512,0.006233777850866318
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,512,0.14296444257100424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,3072,0.14384710788726807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,8192,256,0.012784000072214337
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,256,0.005948444621430502
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,8192,128,0.003549333247873518
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,2048,0.007622222105662028
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,8192,128,0.01239999963177575
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,8192,64,0.0031946665710873077
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,8192,128,0.005919999960396025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,8192,32,0.003528000166018804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,8192,64,0.012459555433856117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,8192,768,0.005590222362014983
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,8192,32,0.01240355521440506
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,7168,65536,0.16936800214979383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,8192,4096,0.02366399930583106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,7168,16384,0.04574133290184868
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,7168,65536,0.22035111321343315
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,8192,256,0.003842666745185852
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,7168,16384,0.0642684433195326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,65536,0.07754488786061604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,7168,12288,0.036313778824276395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,16384,0.14069067107306585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,16384,0.02492977844344245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,7168,12288,0.05082311232884725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,7168,10240,0.030800888935724895
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,12288,0.14514133665296766
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,12288,0.019360888335439894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,128,0.13446222411261663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,65536,0.27852622667948407
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,7168,10240,0.04394222299257914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,10240,0.017136888371573556
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,10240,0.14301244417826334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,7168,8192,0.036826666858461164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,7168,7168,0.022553778356975977
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,8192,0.014999111493428549
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,7168,7168,0.031934221585591636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,7168,0.013978666729397245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,7168,6144,0.019941333267423842
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,7168,0.1452960040834215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,7168,6144,0.02895911203490363
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,6144,0.14711911148495144
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,7168,5120,0.017017778423097398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,5120,0.13900355497996011
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,7168,5120,0.025827555192841425
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,256,0.14356799920399985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,7168,4096,0.014799111419253878
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,5120,0.01200977795653873
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,7168,4096,0.023403555154800415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,4096,0.14808711740705702
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,7168,3584,0.013266666895813413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,4096,0.01038933296998342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,7168,8192,0.026245334082179602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,7168,3072,0.011815110842386881
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,7168,3584,0.021670222282409668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,3584,0.009317333499590555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,3584,0.14052532778845891
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,8192,1024,0.14432266023423937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,7168,2560,0.010318222145239512
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,8192,0.14558045069376627
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,3072,0.008604444563388824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,3072,0.14898578325907388
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,7168,2048,0.009165333376990424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,2560,0.00795111142926746
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,2560,0.14543288283877903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,7168,2560,0.01920000049802992
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,7168,1536,0.00775111135509279
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,7168,2048,0.01756266587310367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,2048,0.14019733005099827
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,2048,0.0075937774446275495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,7168,1024,0.005593777944644292
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,7168,1536,0.015697777271270752
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,1536,0.007300444775157505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,1536,0.14203911357455784
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,7168,768,0.005164444446563721
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,7168,1024,0.014152889450391134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,1024,0.006632888896597757
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,1024,0.13997066020965576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,7168,512,0.004368888835112254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,768,0.14574755562676325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,768,0.0062871111763848206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,7168,256,0.003847111016511917
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,7168,512,0.013407111167907715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,512,0.006248000181383557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,7168,256,0.012727999852763282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,256,0.13888000117407903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,256,0.0059279998143513995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,7168,128,0.003543111185232798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,7168,128,0.012375999655988483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,128,0.005896889087226655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,128,0.13894932799869114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,7168,64,0.0031964443624019623
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,7168,32,0.0035235554807715942
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,7168,64,0.01239377756913503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,7168,32,0.012422222230169507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,6144,65536,0.15319377846188015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,7168,6144,0.012991111311647626
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,6144,65536,0.18393954965803358
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,6144,16384,0.042097777128219604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,65536,0.07527022229300605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,7168,768,0.013779555757840475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,6144,16384,0.05494222376081678
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,6144,12288,0.0333591103553772
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,16384,0.02314844396379259
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,16384,0.1455671124988132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,6144,10240,0.028015110227796767
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,6144,12288,0.04361777835422092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,12288,0.018577777677112155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,12288,0.13960888650682238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,6144,10240,0.03677155574162801
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,10240,0.016844444804721408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,65536,0.26504267586602104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,6144,8192,0.03189333279927572
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,6144,7168,0.021336000826623704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,8192,0.014648889501889547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,7168,3072,0.02058577703105079
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,6144,7168,0.02755466600259145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,7168,0.14088443915049234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,7168,0.01367377738157908
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,6144,6144,0.025246222813924152
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,6144,0.012716444830099741
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,6144,0.1410977840423584
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,6144,5120,0.015973332855436537
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,6144,5120,0.02288444505797492
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,5120,0.14259644349416098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,6144,4096,0.013783999615245394
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,5120,0.011527111132939657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,6144,4096,0.02108088963561588
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,7168,512,0.14429867267608643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,4096,0.14356356196933323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,6144,3584,0.01276177747382058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,6144,8192,0.02422399984465705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,4096,0.009676444033781687
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,10240,0.14494044250912136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,6144,3584,0.019700444406933255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,3584,0.14206222693125406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,3584,0.009185777770148383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,6144,2560,0.011036444041464063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,6144,3072,0.01891288823551602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,3072,0.1420151127709283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,6144,6144,0.018460444278187223
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,6144,2048,0.010273777776294285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,2560,0.008216888540320927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,2560,0.14382933245764837
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,6144,2048,0.015855110353893705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,6144,1536,0.008290666672918532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,2048,0.007638221813572778
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,2048,0.1416755517323812
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,6144,1536,0.014512000812424554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,1536,0.007246221933099959
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,1536,0.14650132921006945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,6144,1024,0.013789333403110504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,1024,0.1449822187423706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,6144,768,0.005225777626037598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,1024,0.006614222294754452
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,6144,3072,0.01109777804878023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,6144,768,0.013431111143694984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,6144,512,0.00489688871635331
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,768,0.14393422338697645
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,768,0.00628266649113761
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,3072,0.008619555168681675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,6144,512,0.01310755560795466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,6144,256,0.0038533334930737815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,512,0.006237333433495627
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,6144,2560,0.017183999220530193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,6144,128,0.0037671112351947357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,6144,256,0.012720000412729053
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,256,0.005949333310127258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,6144,128,0.012103999654452005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,6144,64,0.0031999999450312722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,6144,32,0.0035004446076022256
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,6144,128,0.005938666562239329
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,128,0.1338248915142483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,6144,32,0.01238666640387641
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,6144,1024,0.006294222341643439
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,5120,65536,0.14099021752675375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,5120,65536,0.17824888229370117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,65536,0.07279644409815471
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,5120,16384,0.053583999474843345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,16384,0.022276444567574397
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,16384,0.14214666684468588
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,8192,0.14320354991488987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,5120,12288,0.03573688864707947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,65536,0.26234666506449383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,512,0.14175021648406982
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,5120,12288,0.04119999872313605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,12288,0.018174222773975797
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,12288,0.14469956027136907
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,6144,256,0.14582133293151855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,5120,8192,0.025560888979170058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,5120,10240,0.036497778362698026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,10240,0.016201777590645682
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,10240,0.1404106616973877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,5120,7168,0.022321777211295232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,5120,8192,0.031033777528338965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,8192,0.014283556077215405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,8192,0.14478222529093424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,5120,6144,0.019847111569510568
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,6144,64,0.012063999970753988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,5120,7168,0.027112000518374976
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,5120,5120,0.01677955521477593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,7168,0.013190222283204397
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,5120,6144,0.024720888998773363
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,5120,16384,0.04541511005825467
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,6144,0.012062222593360476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,6144,0.14170666535695395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,5120,5120,0.02277333372169071
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,5120,0.010943111446168689
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,5120,3584,0.012774222426944308
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,4096,0.14135289192199707
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,4096,0.009305777649084726
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,5120,3584,0.019497777024904888
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,3584,0.00869066682126787
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,5120,3072,0.011510222322411008
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,5120,3072,0.018171555466122098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,3072,0.14472177293565539
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,5120,10240,0.030155556069480047
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,5120,2560,0.010264000131024255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,3072,0.008295111358165741
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,5120,2560,0.016887111796273124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,2560,0.14585155910915798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,5120,2048,0.008932444784376357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,2560,0.007930667036109501
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,5120,2048,0.015464888678656684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,2048,0.13971288998921713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,7168,0.14168177710639104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,5120,1536,0.007304888632562425
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,2048,0.007332444190979004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,5120,4096,0.01425155500570933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,5120,1536,0.014477334088749357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,5120,0.14500621954600015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,5120,1024,0.005605333381228977
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,1536,0.14047555128733316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,1536,0.006933333145247565
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,5120,4096,0.020623111062579684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,5120,768,0.00489688871635331
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,5120,1024,0.013748444616794586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,1024,0.143321779039171
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,5120,768,0.013377777404255338
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,768,0.00627377794848548
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,768,0.14683022763994005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,3584,0.14157422383626303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,5120,256,0.0038853333228164246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,5120,512,0.01311377767059538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,512,0.14484532674153647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,5120,256,0.012453333371215396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,5120,128,0.0035004446076022256
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,256,0.14643200238545737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,256,0.005893333504597346
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,5120,64,0.0031795555518733132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,5120,128,0.012095111111799875
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,5120,128,0.1365440024269952
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,5120,32,0.0031866667171319327
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,5120,64,0.012427555190192329
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,5120,32,0.01240888900227017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,4096,65536,0.1010871132214864
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,4096,65536,0.1723466714223226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,65536,0.07083821958965726
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,4096,16384,0.03360177742110358
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,1024,0.006620444357395172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,4096,16384,0.05081689026620653
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,5120,512,0.0041759999261962045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,16384,0.14488266574011907
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,65536,0.2598337862226698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,16384,0.021363554729355708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,512,0.00620977787507905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,4096,10240,0.01999022232161628
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,4096,12288,0.038839999172422625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,12288,0.1453617811203003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,12288,0.01754488878779941
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,4096,8192,0.0170817772547404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,10240,0.01566488875283135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,4096,10240,0.03477155499988132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,10240,0.14291644096374512
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,4096,7168,0.015215999550289579
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,5120,128,0.005903111149867375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,4096,8192,0.03036710951063368
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,8192,0.013737777868906656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,8192,0.14719377623664007
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,7168,0.012568888564904531
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,7168,0.13966666327582464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,4096,5120,0.01143200033240848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,6144,0.14482133918338352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,4096,5120,0.022508444057570562
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,5120,0.14529777897728816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,4096,4096,0.010695999695195092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,5120,0.010371555884679159
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,4096,4096,0.020260444945759244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,4096,12288,0.027434666951497395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,4096,3584,0.009008000294367472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,4096,0.0090515555606948
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,4096,3584,0.018551111221313477
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,3584,0.14015911685095891
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,4096,3072,0.008678221868144141
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,3584,0.008722667064931657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,4096,6144,0.013388444152143268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,3072,0.008279999925030602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,3072,0.14048177666134304
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,4096,2560,0.007811555431948767
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,4096,7168,0.026202667090627883
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,4096,2048,0.006871999965773688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,4096,6144,0.024294222394625347
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,4096,2560,0.01650399963061015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,2560,0.007676444119877285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,6144,0.011398221883508893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,2560,0.14107555813259548
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,4096,1536,0.0052871112194326185
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,4096,2048,0.01517599986659156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,2048,0.0074311114019817775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,4096,1024,0.00423200014564726
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,2048,0.14031999640994602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,4096,1536,0.014186667071448432
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,1536,0.1404906643761529
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,4096,768,0.004096888833575779
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,4096,1024,0.01349955548842748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,1024,0.00656355544924736
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,1024,0.14325066407521567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,4096,512,0.0038444445365005066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,4096,768,0.013109332985348172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,768,0.14726311630672878
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,768,0.006278222219811545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,4096,0.13971288998921713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,4096,256,0.003528888854715559
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,4096,512,0.012785777449607849
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,512,0.005956444475385878
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,4096,128,0.0031662223239739737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,4096,256,0.012466666599114736
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,256,0.005939555664857228
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,4096,128,0.011868444581826528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,4096,64,0.0031644445326593188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,128,0.005943111247486538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,128,0.13604177369011775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,4096,32,0.0028595555987623003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,4096,3072,0.01789155602455139
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,4096,64,0.012105777859687805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,4096,32,0.012421333127551608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3584,65536,0.09347644117143418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3584,16384,0.03119200136926439
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3584,65536,0.1662951045566135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,65536,0.0698453320397271
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3584,16384,0.04952977763281929
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,4096,1536,0.0069146665434042616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3584,12288,0.024864888853496973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,16384,0.020940444535679285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,16384,0.14356799920399985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3584,10240,0.02208799951606327
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3584,12288,0.037815110550986394
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,65536,0.25904621018303764
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,12288,0.017298666967286002
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3584,8192,0.016197333733240765
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3584,10240,0.03397866752412584
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,512,0.13978666729397243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,10240,0.14681243896484375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,10240,0.015587555037604438
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3584,7168,0.014230221509933472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,4096,256,0.13950488302442762
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,8192,0.013412444127930535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,8192,0.14310044712490505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3584,7168,0.02609599961174859
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3584,6144,0.013070222404268054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,7168,0.14464799563090006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,7168,0.012292444705963135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3584,5120,0.01144977741771274
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,6144,0.011193777951929303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,6144,0.14130400286780462
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3584,5120,0.022064000368118286
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3584,4096,0.011728888584507836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,5120,0.009988444546858469
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3584,4096,0.019945777124828763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,4096,0.14280711279975042
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,4096,0.009007111191749573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3584,3584,0.010704889065689512
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3584,3584,0.01819022165404426
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,3584,0.008389333056079017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,3584,0.1412995523876614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3584,3072,0.008271999657154083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3584,3072,0.016846223009957206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,3072,0.008276444342401292
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,12288,0.1432622273763021
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,3072,0.13971111509535047
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3584,2560,0.007616000043021307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3584,2048,0.006334222025341458
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,2560,0.007939555578761632
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,2560,0.14252621597713894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3584,8192,0.029829333225886028
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3584,1536,0.0052737775776121355
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,2048,0.007303111255168915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3584,1536,0.014470222923490735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,1536,0.1430035564634535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3584,1024,0.004212444441186057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,1536,0.006996444529957241
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3584,1024,0.01347555551264021
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,1024,0.14769689242045084
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,1024,0.006646222124497096
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3584,768,0.0131217779384719
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,768,0.14668888515896267
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3584,512,0.0035137778355015647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3584,512,0.01277511070171992
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3584,2048,0.015136889285511441
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,2048,0.14015021589067247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,512,0.1456880039638943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,512,0.005939555664857228
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3584,256,0.0034835553831524322
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3584,6144,0.024248000648286607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3584,256,0.012072888513406118
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,256,0.00590133335855272
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,256,0.14224178261227077
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,5120,0.14981688393486872
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3584,64,0.0031431110368834603
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3584,128,0.011727111207114326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,128,0.0058906666106647914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3584,32,0.003181333343187968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3584,768,0.003871110992299186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3584,64,0.01205244412024816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3584,32,0.012359999948077731
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3072,65536,0.13123022185431585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,65536,0.0684915582338969
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,3584,768,0.006287999865081575
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3072,16384,0.02665511104795668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3584,2560,0.016158221496476066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3072,16384,0.03944977786805894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,16384,0.1472302277882894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3072,12288,0.02258844508065118
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3072,12288,0.031488888793521456
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,12288,0.14450844128926596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3072,10240,0.021977777282396953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,12288,0.01720888912677765
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3584,128,0.003145777723855443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3072,10240,0.027430221438407898
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,10240,0.015101333459218344
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3072,8192,0.018840889135996502
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,3584,128,0.13246933619181314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3072,8192,0.025017778078715008
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3072,65536,0.07736800114313762
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,8192,0.14530755413903132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3072,7168,0.016524444023768108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3072,7168,0.02198933396074507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,7168,0.011922666596041786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,7168,0.14644532733493382
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3072,6144,0.015438222222858004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3072,6144,0.02055911057525211
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,6144,0.010716444088353051
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3072,5120,0.012696888711717395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,65536,0.2571546766493055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3072,5120,0.019175110591782462
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3072,4096,0.011723555624485016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,5120,0.010008888940016428
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3072,4096,0.01685333251953125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,4096,0.14467822180853948
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3072,3584,0.010703110860453712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,4096,0.008777778181764815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,10240,0.13999288611941868
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3072,3584,0.016146666473812528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,3584,0.008636444807052612
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3072,3072,0.009639999932712978
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3072,3072,0.0151235560576121
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,3072,0.15145066049363878
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3072,2560,0.008618666893906063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,3072,0.00831911133395301
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3072,2560,0.014441778262456259
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3072,2048,0.007660444411966536
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,6144,0.14563111464182535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,2560,0.007924444145626491
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,16384,0.020830222302012973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3072,1536,0.005884444548024072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3072,2048,0.013757333159446716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,2048,0.007368889119890001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,5120,0.14395556184980604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,2048,0.14546133412255183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3072,1024,0.00488711107108328
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3072,1536,0.013410666750537025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,1536,0.0069671107663048645
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,1536,0.1496773295932346
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,1024,0.1445280048582289
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,1024,0.00656444455186526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3072,768,0.012732444538010491
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3072,512,0.0038444445365005066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,768,0.006246222390068903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,768,0.14518577522701687
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,3584,0.14771289295620388
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3072,512,0.012376888758606382
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,512,0.005970666805903117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,512,0.14355466100904676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3072,256,0.011762667033407422
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,256,0.005923555543025334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,256,0.14502310752868652
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3072,128,0.011717333561844297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,128,0.13737332820892334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3072,64,0.0029093333416514923
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,128,0.0058959999846087555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3072,32,0.0031866667171319327
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3072,64,0.012083555261294046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2560,65536,0.06868888934453328
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2560,65536,0.1281342241499159
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,3072,8192,0.013112000293201871
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3072,256,0.003524444583389494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2560,16384,0.023737778266270954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,65536,0.06734844711091784
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3072,128,0.003177777760558658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2560,16384,0.038691557115978666
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,65536,0.25649245580037433
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2560,12288,0.019674667053752475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,3072,2560,0.1417893303765191
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,16384,0.02051999999417199
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,16384,0.1383404466840956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2560,10240,0.01793155570824941
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,12288,0.016607999801635742
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2560,12288,0.030029333300060693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,12288,0.14626222186618382
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2560,10240,0.027432888746261597
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2560,8192,0.016511110795868766
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,10240,0.14647200372483996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,10240,0.015103111664454142
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2560,7168,0.01497866710027059
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2560,8192,0.02492444382773505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,8192,0.012440889245933957
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3072,32,0.011762667033407422
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,8192,0.13839822345309788
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2560,6144,0.01275111072593265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2560,7168,0.021831111775504217
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,7168,0.011394666300879585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,7168,0.14543999565972224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2560,5120,0.012648888760142855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2560,6144,0.02027555637889438
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,6144,0.010640888578361936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,6144,0.13920888635847303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2560,4096,0.010328888893127441
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2560,5120,0.018100443813535903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,5120,0.009855111440022787
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,5120,0.1452488899230957
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2560,4096,0.016844444804721408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2560,3584,0.009293333523803288
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,4096,0.14581778314378527
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,4096,0.008990222381220924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2560,3072,0.008570666942331526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2560,3584,0.015671110815472074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,3584,0.008602666358153025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,3584,0.14023644394344753
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2560,2560,0.007357333269384172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2560,3072,0.014803555276658801
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,3072,0.14520000086890325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2560,2048,0.006614222294754452
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,2560,0.14383555783165827
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,2560,0.007638221813572778
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2560,2048,0.013849777479966482
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,2048,0.007311111523045435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,2048,0.1441484424802992
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2560,1536,0.013099555340078143
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,3072,768,0.004512000001139111
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,1536,0.14335466755761042
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2560,1024,0.004509333521127701
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,1536,0.006949333681000604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2560,1024,0.012776888906955719
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2560,768,0.003908444609906939
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,1024,0.006586666736337874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2560,768,0.012429333395428128
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,768,0.1492906676398383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2560,512,0.0035422220826148987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,768,0.006246222390068903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2560,512,0.012071111136012606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,3072,0.008270222279760573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2560,2560,0.014464888307783337
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2560,256,0.0034933334423436057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,512,0.005917333480384614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2560,256,0.011864888999197217
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2560,1536,0.005515555540720622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2560,128,0.0031893334041039148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2560,128,0.011707555916574268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,128,0.0058853332367208265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,128,0.13555644618140325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,3072,1024,0.013048888908492194
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2560,64,0.003181333343187968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2560,32,0.003168888803985384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2048,65536,0.05483288897408379
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2560,64,0.011671110987663269
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2560,32,0.011773332953453064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2048,16384,0.02015022271209293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2048,65536,0.12557066811455622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,65536,0.06631111436420016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2048,16384,0.03710666630003188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2048,12288,0.016490666402710807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,16384,0.14627199702792698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,16384,0.020232889387342665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,65536,0.2557786570654975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2048,10240,0.015673778123325773
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2048,12288,0.02974577744801839
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,12288,0.016182222300105624
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,12288,0.14512622356414795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,512,0.14296089278327093
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2048,8192,0.01427288850148519
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2048,10240,0.02700800034734938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,10240,0.013971555564138623
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,10240,0.1414008935292562
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2048,7168,0.013434666726324292
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2048,8192,0.02425777746571435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,256,0.14495910538567439
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,8192,0.1468800041410658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,8192,0.01203555530971951
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2048,6144,0.012774222426944308
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,7168,0.011129777464601727
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,7168,0.14005333847469753
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2048,6144,0.019613333874278598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,6144,0.01073066641887029
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,6144,0.1417377789815267
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2048,4096,0.009280888570679558
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2048,5120,0.017895110779338412
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,5120,0.009760888914267222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,5120,0.14504178365071616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,2560,1024,0.14056177934010824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2048,4096,0.016190222567982145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,4096,0.14020000563727483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2048,3584,0.015497777197096082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2048,3072,0.0075795559419526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,3584,0.008541333178679148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2048,3072,0.014800889624489678
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,3072,0.13893066512213811
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2048,2560,0.006455111006895701
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,3072,0.007988444632954067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2048,2560,0.01408088869518704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,2560,0.14024089442359075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,2560,256,0.005886222339338726
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2048,7168,0.02130399975511763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2048,2048,0.005535111245181825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,2560,0.007636444436179266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2048,5120,0.010473777850468954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,2048,0.007304000357786815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2048,1536,0.005215999980767568
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,2048,0.14095377922058105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2048,1536,0.013056000073750814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,1536,0.0069466663731469055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2048,1024,0.0041706665522522396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2048,3584,0.00795555528667238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,4096,0.009032888544930352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2048,768,0.0039022221333450745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2048,1024,0.012718222207493253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,1024,0.14057332939571804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,3584,0.14151644706726074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2048,512,0.003504000190231535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,768,0.006266666783226862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2048,512,0.012062222593360476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,512,0.14151377148098415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,512,0.005943999936183293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2048,256,0.011776000261306763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2048,2048,0.013565333353148567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,256,0.1422648827234904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2048,128,0.002867555452717675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,256,0.00592533333433999
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,1536,0.1446062193976508
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2048,128,0.011603555745548673
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2048,64,0.0028400001012616684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,128,0.13878133561876085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,128,0.005791999813583162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,2048,1024,0.006606222026877933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2048,32,0.0028435554769304064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2048,64,0.011426666544543372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1536,65536,0.04524533285035027
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2048,32,0.011761777930789523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,2048,768,0.14286577701568604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1536,16384,0.016585777203241985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1536,65536,0.12188443872663711
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,65536,0.06400089131461249
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1536,16384,0.03620355659061008
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1536,12288,0.01421688828203413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,16384,0.1424071126514011
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,2048,256,0.003175111073586676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,65536,0.25539199511210126
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1536,10240,0.014766222900814481
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1536,12288,0.029475556479560003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,12288,0.015522667103343539
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,12288,0.14495733049180773
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1536,8192,0.014868444866604276
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1536,10240,0.02648711038960351
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,10240,0.1411964495976766
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,10240,0.013713777893119387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1536,7168,0.012528888881206512
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1536,8192,0.023994666006830003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,8192,0.011752888560295105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,8192,0.14222578207651773
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1536,6144,0.011356444822417365
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1536,7168,0.020966221888860066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,7168,0.011023999916182624
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,7168,0.14005422592163086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1536,5120,0.010072000324726105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1536,6144,0.019389332996474374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,6144,0.14022489388783774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,6144,0.010663111176755695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1536,4096,0.008887999587588841
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1536,5120,0.017758222089873422
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,5120,0.009665778113736046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,5120,0.14574044280582005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1536,3584,0.007992000215583378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1536,4096,0.01612622208065457
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,2048,768,0.01238844460911221
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,4096,0.008990222381220924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,4096,0.14115466011895075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1536,3072,0.007105777661005656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1536,3584,0.01519733336236742
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,3584,0.008391111261314815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,3584,0.14272799756791857
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1536,3072,0.014660444524553088
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,3072,0.008011555506123437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,3072,0.14077511098649767
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1536,2048,0.005534222142563925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1536,2560,0.014143110977278816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,16384,0.019296889503796894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,2560,0.007642666498819987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1536,1536,0.004832889056868023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1536,2048,0.013625777430004544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,2048,0.13997688558366564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1536,1024,0.004181333300140169
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1536,1536,0.013064888616402945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,1536,0.006940444310506185
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,1536,0.139847993850708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1536,768,0.003839111162556542
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1536,1024,0.01272266689274046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,1024,0.006636444479227066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1536,512,0.0035253332720862497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1536,768,0.012423999607563019
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,768,0.14147999551561144
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,768,0.006265777680608962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1536,256,0.00319377767542998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1536,512,0.012071111136012606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,512,0.005956444475385878
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1536,256,0.01203644441233741
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,256,0.14757688840230307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,256,0.0058764442801475525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1536,128,0.002921777880854077
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1536,2560,0.006575999988449945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1536,128,0.011735999749766456
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,128,0.13893777794308132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1536,64,0.0027955555253558685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1536,32,0.0027919999427265594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,128,0.0059368887709246735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,2560,0.1422817839516534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1536,64,0.011363555159833698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1024,65536,0.033951110310024686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1536,32,0.011723555624485016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1024,16384,0.013621333572599622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,65536,0.06179022126727634
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1024,65536,0.11777955955929226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,1024,0.1424915525648329
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1024,16384,0.03540800015131632
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,65536,0.2561093436347114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1024,12288,0.01460355520248413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,16384,0.1456880039638943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,16384,0.01787644459141625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1024,10240,0.013408888545301227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1024,12288,0.029120001527998183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,12288,0.01479644411140018
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,12288,0.1490373346540663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,1536,512,0.14019821749793157
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1024,10240,0.025867554876539443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,10240,0.013377777404255338
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,10240,0.14467822180853948
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1024,7168,0.011665778027640449
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1024,8192,0.022320000661744013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,8192,0.011740444435013665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1024,6144,0.010663111176755695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1024,7168,0.019965334071053397
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,7168,0.011046222514576383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,7168,0.14055021603902182
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1024,5120,0.009711999860074785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1024,6144,0.018555555078718398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,6144,0.010427555276287926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,6144,0.14455466800265843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1024,4096,0.007910221815109253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1024,5120,0.01719466679626041
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,5120,0.009680888719028896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,1536,2048,0.007312888900438945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1024,4096,0.016169778174824186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1024,3584,0.01038577738735411
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,4096,0.008656889200210571
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,4096,0.14133155345916748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1024,3072,0.00980977796845966
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1024,3584,0.015152000718646579
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,3584,0.008303110798199972
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,3584,0.14744177129533556
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1024,2560,0.00830666638082928
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1024,3072,0.014537778165605334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,3072,0.007967111137178209
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,3072,0.144122666782803
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1024,2048,0.0074728892909155945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1024,2560,0.013787555197874704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1024,8192,0.012337778177526263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,2560,0.14191110928853354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,2560,0.007609777980380588
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1024,1536,0.00489688871635331
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1024,2048,0.013574221895800697
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,2048,0.007288888924651676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,2048,0.14542578326331243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,8192,0.1389520035849677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1024,1024,0.0042408886882993905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,1536,0.14100621806250677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,1536,0.006952889263629913
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1024,1024,0.01242844429281023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,1024,0.14077156119876438
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1024,768,0.01240888900227017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1024,512,0.003524444583389494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,768,0.006269333263238271
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,768,0.14671111106872559
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,5120,0.14471199777391222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1024,512,0.012060444388124677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1024,256,0.0034871109657817413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,512,0.005936000082227919
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,512,0.13957244820064968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1024,128,0.003140444349911478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1024,256,0.012084444363911947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,256,0.005821333163314396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,256,0.1438373327255249
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1024,64,0.002862222285734283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1024,32,0.0028693332440323303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1024,128,0.011408888631396823
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,1024,128,0.13288888666364881
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1024,64,0.011771555576059552
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,768,65536,0.026799110902680293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1024,32,0.011363555159833698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,768,16384,0.014300444059901767
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,768,65536,0.11682577927907307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,65536,0.061258667045169406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,768,16384,0.0351582235760159
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,1024,1536,0.012784889174832238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,768,12288,0.0125608891248703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,16384,0.017869333426157635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,1024,768,0.003871110992299186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,1024,0.006592000110281839
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,768,12288,0.028202666176689997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,768,10240,0.013075555364290873
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,12288,0.14614933066897923
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,768,10240,0.025081778566042583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,10240,0.013448889056841532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,10240,0.14241333802541098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,768,8192,0.022745778163274128
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,8192,0.14196266068352595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,768,7168,0.01072177787621816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,8192,0.011728888584507836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,1024,128,0.005928888916969299
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,768,7168,0.019898666275872123
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,768,6144,0.009703999592198266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,7168,0.14675377474890813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,7168,0.011049778097205691
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,768,6144,0.018541332748201158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,6144,0.010388444695207808
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,6144,0.14524266454908583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,768,5120,0.009152000149091085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,65536,0.2558702097998725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,768,4096,0.008000888758235509
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,16384,0.1417306661605835
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,768,5120,0.017216889394654166
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,5120,0.14590666029188368
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,768,3584,0.00793333351612091
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,12288,0.0148062225845125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,768,4096,0.01609422266483307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,4096,0.1423084470960829
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,768,3072,0.00655733338660664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,768,3584,0.014904888139830695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,3584,0.00832622249921163
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,768,8192,0.011376889215575324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,3584,0.13971377743615046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,768,2560,0.005912888795137405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,768,3072,0.014595554934607612
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,3072,0.008007999923494127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,3072,0.14604711532592773
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,768,2048,0.0052586665583981415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,768,2560,0.01408088869518704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,2560,0.1481377813551161
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,2560,0.0076453329788313965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,768,1536,0.00489066665371259
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,768,2048,0.013408888545301227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,2048,0.007296889192528195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,2048,0.14462133248647055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,768,1024,0.0038488888078265717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,768,1536,0.013063111239009433
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,1536,0.0069146665434042616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,1536,0.14691821734110513
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,768,768,0.003543111185232798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,768,1024,0.012734221915404001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,1024,0.13970577716827393
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,1024,0.006600889066855113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,768,512,0.003535110917356279
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,768,768,0.012091555529170565
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,768,0.006226666685607698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,768,0.14546133412255183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,768,512,0.012089778151777057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,512,0.00591111100382275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,512,0.14011111524369982
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,5120,0.009655999640623728
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,768,128,0.0028728888266616394
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,768,256,0.011707555916574268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,256,0.005919111271699269
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,256,0.14083288775549993
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,768,64,0.0028479999552170434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,4096,0.009013333254390294
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,768,128,0.011424889167149862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,768,32,0.0028417776856157514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,768,128,0.13185866673787436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,768,128,0.0058853332367208265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,512,65536,0.027831110689375136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,768,64,0.011693333586057028
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,768,32,0.011724444727102915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,512,16384,0.016492444607946608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,65536,0.06036800146102905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,512,65536,0.11555733945634629
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,512,12288,0.011720000041855706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,65536,0.25476622581481934
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,16384,0.017814222309324477
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,16384,0.14436533715989855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,512,10240,0.01035911093155543
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,512,12288,0.027830223242441814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,12288,0.14684089024861655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,12288,0.01482311056719886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,512,8192,0.008971555365456475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,512,10240,0.024620445238219366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,10240,0.14784888426462808
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,10240,0.013412444127930535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,512,8192,0.021977777282396953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,8192,0.011706666813956367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,8192,0.1400008863872952
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,512,6144,0.008355555435021719
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,512,7168,0.01961866683430142
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,7168,0.011165333290894827
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,7168,0.14043733808729383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,512,5120,0.007648888561460707
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,512,6144,0.01866222255759769
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,768,256,0.0031519999934567344
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,6144,0.15024356047312418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,512,4096,0.007609777980380588
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,512,5120,0.017166222135225933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,5120,0.1403991116417779
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,512,4096,0.015800888339678448
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,4096,0.1419022215737237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,512,3584,0.007250666618347168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,512,3584,0.01512977812025282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,3584,0.14619199434916177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,512,3072,0.006648889018429651
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,3584,0.008354666332403818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,512,3072,0.014447111222479077
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,3072,0.14136977990468344
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,512,2560,0.005903999838564131
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,512,7168,0.008648888932334052
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,512,2560,0.013821333646774292
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,2560,0.14660355779859754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,512,2048,0.005212444398138259
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,2560,0.007639110916190677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,512,2048,0.01349955548842748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,5120,0.009651555783218807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,512,1536,0.004575110971927643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,2048,0.14205333921644422
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,2048,0.007319110963079665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,4096,0.00868444475862715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,512,1024,0.0038284444146686127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,512,1536,0.01311555587583118
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,1536,0.14152799712287054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,1536,0.006931555353932911
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,512,768,0.0035439998739295532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,512,1024,0.01275111072593265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,1024,0.006593777901596493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,512,16384,0.034902221626705594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,512,768,0.012423111332787408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,512,512,0.0034871109657817413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,768,0.00628977765639623
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,768,0.1397848924001058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,512,256,0.0031600000543726813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,512,512,0.012079111403889127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,512,0.005966222120655908
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,512,0.14249689049190944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,3072,0.008009778128729926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,512,128,0.00286666676402092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,512,256,0.011697777443461947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,512,64,0.0028088889602157804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,256,0.005927111125654644
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,512,32,0.0028355556229750314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,128,0.005889777921968036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,512,128,0.01167377746767468
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,128,0.13799911075168186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,256,65536,0.019135110908084445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,512,64,0.011351111034552256
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,256,16384,0.009352000223265754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,512,32,0.011713777979214987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,65536,0.05624622106552124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,256,16384,0.03324800067477756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,65536,0.2535333368513319
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,16384,0.14493866761525473
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,256,12288,0.008603555460770925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,16384,0.017842666970358956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,256,10240,0.012395555774370829
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,256,12288,0.027067555321587458
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,12288,0.0147733340660731
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,12288,0.145288003815545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,256,8192,0.010760000182522668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,256,10240,0.024160888459947374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,10240,0.013410666750537025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,10240,0.14423110749986437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,256,7168,0.010112000008424124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,256,8192,0.02163733376397027
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,8192,0.14225067032708064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,8192,0.011688888900809817
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,256,6144,0.007662222617202335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,7168,0.14460444450378418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,7168,0.011020444333553314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,256,0.1404204501046075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,256,6144,0.018204443984561496
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,6144,0.010346666806273991
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,6144,0.14665067195892334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,256,5120,0.016885333591037326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,512,6144,0.010354667074150508
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,256,4096,0.007600000335110559
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,5120,0.009665778113736046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,5120,0.14158666133880615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,256,3584,0.007257777783605788
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,256,65536,0.11266133520338271
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,4096,0.00867733359336853
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,4096,0.14682755205366346
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,256,3072,0.00692622239391009
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,256,3584,0.01479466590616438
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,3584,0.008339555727110969
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,3584,0.1439324484931098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,256,2560,0.005568888866239124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,256,3072,0.014087999860445658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,3072,0.008004444340864817
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,3072,0.14279645019107395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,512,1024,0.14312977261013454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,256,2560,0.014123555686738757
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,2560,0.007679111427730984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,2560,0.14424888292948404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,256,2048,0.013741333451535968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,2048,0.14317333698272705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,256,7168,0.019596444235907663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,2048,0.007270221908887227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,256,1536,0.004238222208287981
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,256,1024,0.0038782221575578055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,1536,0.0069208890199661255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,256,1536,0.013112000293201871
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,1536,0.13937066660987005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,256,1024,0.012715555727481842
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,256,5120,0.00833155545923445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,256,768,0.003527111063400904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,1024,0.006603555546866522
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,256,512,0.003244444520937072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,768,0.006247111078765657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,256,768,0.01237066669596566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,768,0.14100533061557347
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,256,256,0.0028728888266616394
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,256,512,0.01203555530971951
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,512,0.005956444475385878
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,256,4096,0.015855999456511605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,512,0.1394995583428277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,256,128,0.0029164445069101122
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,256,256,0.011758222348160215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,256,0.0058986664646201665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,256,0.14211822880638972
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,256,128,0.0058782220714622075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,128,0.13945244418250188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,256,32,0.002509333400262727
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,128,65536,0.015004444453451367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,256,64,0.011695110963450538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,256,2048,0.004877333425813251
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,256,32,0.011049778097205691
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,128,16384,0.009661333428488838
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,128,65536,0.11223288377126057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,65536,0.052821334865358144
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,128,16384,0.03282577792803446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,65536,0.2560479905870226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,128,12288,0.008343111309740279
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,16384,0.01777866648303138
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,16384,0.13946933216518825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,128,10240,0.009272889130645329
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,128,12288,0.026983999543719824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,12288,0.014783999986118741
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,12288,0.14670222335391575
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,256,1024,0.14445777734120688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,128,10240,0.02425866656833225
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,10240,0.013415999710559845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,128,8192,0.021312889125612047
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,8192,0.13901244269476995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,8192,0.011710222396585675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,256,128,0.011408888631396823
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,128,7168,0.019816888703240287
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,7168,0.14176889260609946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,7168,0.011071110765139261
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,128,6144,0.0181404451529185
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,6144,0.13890133963690862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,6144,0.010344888601038191
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,128,5120,0.006958222223652734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,128,5120,0.01682400041156345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,5120,0.13989867104424372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,128,4096,0.006616000086069107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,5120,0.009704888694816166
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,128,8192,0.00831200016869439
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,128,4096,0.015509333875444202
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,10240,0.14290489090813532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,128,3584,0.006508444332414203
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,4096,0.14222310649024114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,4096,0.008664888640244802
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,128,3072,0.00628266649113761
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,128,7168,0.00794311116139094
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,128,3584,0.015059555570284525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,3584,0.13940533002217612
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,128,2560,0.006263999889294307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,128,3072,0.014112000664075216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,3072,0.008096000386608971
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,128,6144,0.007306666837798224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,128,2048,0.006244444598754247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,128,2560,0.013800000150998434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,2560,0.007332444190979004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,2560,0.14299644364251032
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,128,1536,0.004549333204825719
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,128,2048,0.013449778159459433
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,2048,0.007291555404663086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,2048,0.1424693399005466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,128,1536,0.01311644415060679
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,1536,0.006937777830494775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,128,1024,0.01277599980433782
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,1024,0.006551110910044775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,1024,0.14437954955630833
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,128,768,0.012112889024946423
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,768,0.14553599887424046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,128,512,0.0032497776879204642
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,768,0.006231110956933763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,128,512,0.012100444071822695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,512,0.1388231118520101
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,3584,0.00832711077398724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,128,256,0.0031351111829280853
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,512,0.005984000033802456
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,256,64,0.0028044444819291434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,3072,0.144195556640625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,256,0.1398879951900906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,128,128,0.0028942223224374983
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,256,0.005888888819350137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,128,128,0.011679111255539788
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,160,128,128,0.0058808889653947614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,128,0.14438754982418484
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,128,1024,0.0038284444146686127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,128,32,0.0025351111673646504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,160,128,1536,0.14712444941202799
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,128,64,0.011735111474990845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,128,32,0.011415999796655444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,64,16384,0.009003555609120263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,128,768,0.0035457776652442086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,64,12288,0.008265777594513362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,64,10240,0.007299555672539606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,64,65536,0.11103644635942246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,64,8192,0.007296889192528195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,64,12288,0.026769777139027912
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,64,10240,0.024146666129430134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,64,8192,0.021278222401936848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,64,6144,0.006579555571079254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,64,7168,0.019725332657496136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,64,5120,0.005947555518812603
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,64,6144,0.017901332841979135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,64,4096,0.005575110928879843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,64,5120,0.017183999220530193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,128,256,0.011704888608720569
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,64,3584,0.006365333166387346
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,64,3072,0.006265777680608962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,64,4096,0.01535999940501319
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,64,2560,0.006264888991912206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,64,3584,0.0148062225845125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,64,2048,0.006015111174848344
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,64,3072,0.014454222387737699
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,64,1536,0.004512888689835866
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,64,2560,0.013438222308953604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,64,65536,0.014783110883500842
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,64,1024,0.0038844446341196695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,64,768,0.003535110917356279
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,64,2048,0.013426666458447775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,64,512,0.0032195556494924757
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,64,1536,0.013069333301650153
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,64,1024,0.012320888539155325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,64,256,0.0028684443483750024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,64,128,0.0028026666906144884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,64,512,0.011767110890812345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,64,256,0.011709333293967776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,64,64,0.00281866660548581
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,64,7168,0.0069022224181228215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,64,128,0.011406222151385413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,32,65536,0.013448889056841532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,64,64,0.011354666617181567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,64,32,0.011428444749779172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,32,12288,0.007631111476156447
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,32,65536,0.11082666450076634
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,32,16384,0.03254222207599216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,32,10240,0.007295110987292395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,32,8192,0.0069617778062820435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,32,12288,0.026751110951105755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,32,7168,0.006646222124497096
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,32,10240,0.023997333314683702
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,32,6144,0.006414222220579783
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,32,7168,0.019899555378490023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,128,64,0.0028026666906144884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,32,6144,0.018184888694021437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,32,4096,0.0058986664646201665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,32,5120,0.01716888944307963
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,32,3584,0.0064497776329517365
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,32,4096,0.01574133336544037
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,32,3072,0.005952888892756567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,64,768,0.011750222080283694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,32,2560,0.006264888991912206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,64,16384,0.032688889238569475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,32,2048,0.005924444645643234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,64,32,0.0028168888141711554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,32,3072,0.01423111061255137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,32,1536,0.004202666795916027
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,32,2560,0.013383111192120446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,32,1024,0.003842666745185852
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,32,16384,0.00865955568022198
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,32,1536,0.012744888663291931
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,32,512,0.003149333306484752
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,32,1024,0.012050666742854647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,32,256,0.0029048888633648553
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,32,512,0.011770666473441653
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,32,128,0.002830222249031067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,32,256,0.011101333631409539
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,32,64,0.0028533333291610083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,32,8192,0.02160088883505927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,32,128,0.011361777782440186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,32,64,0.010709332923094431
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,32,5120,0.006263999889294307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,32,32,0.011217777927716574
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,65536,16384,0.22211821873982748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,32,3584,0.0148044443792767
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,65536,16384,0.3758684529198541
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,32,768,0.0035048888789282907
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,65536,12288,0.28083109855651855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,32,768,0.012152888708644442
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,65536,12288,0.16767111089494494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,16384,0.13743466801113552
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,65536,10240,0.23763465881347656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,65536,10240,0.1435706615447998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,12288,0.1049235529369778
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,65536,8192,0.19484621948666045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,10240,0.08702488740285237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,160,32,2048,0.013400889105266996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,65536,8192,0.11903644932640924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,16384,0.3286497857835558
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,12288,0.25659200880262584
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,65536,7168,0.1708373361163669
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,65536,7168,0.10361511177486843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,8192,0.07016266716851129
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,10240,0.22253065639071992
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,160,32,32,0.0028008888992998335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,65536,6144,0.15076888932122126
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,65536,6144,0.08988444672690497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,7168,0.06377600298987494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,8192,0.1866391102472941
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,6144,0.0557617810037401
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,65536,5120,0.07680533329645793
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,65536,4096,0.10062666734059651
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,7168,0.17161333560943604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,5120,0.04723555511898465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,65536,4096,0.06449777550167508
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,65536,3584,0.08858933051427205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,4096,0.039303110705481634
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,65536,3584,0.058528886901007764
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,4096,0.14401422606574163
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,65536,3072,0.0773600008752611
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,3584,0.03515555461247762
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,3584,0.14563022719489202
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,65536,3072,0.0520977775255839
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,3072,0.03147822287347581
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,65536,2560,0.06486222479078504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,65536,5120,0.126367105378045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,3072,0.14248977767096624
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,65536,2560,0.046559111939536206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,2560,0.027777777777777776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,2560,0.14194399780697292
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,65536,2048,0.0414737794134352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,2048,0.024076445235146418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,65536,1536,0.0413680010371738
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,2048,0.14413866731855604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,6144,0.15377067195044622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,65536,1024,0.030432889858881634
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,65536,1536,0.0343431101904975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,1536,0.019288889235920377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,5120,0.14513511127895778
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,65536,768,0.022907555103302002
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,1024,0.01551555593808492
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,65536,1024,0.02845422095722622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,1024,0.14063910643259683
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,65536,768,0.023795555035273235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,768,0.014091556270917257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,768,0.14447111553615993
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,65536,256,0.010676444404655032
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,65536,512,0.020598222812016804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,512,0.01313866674900055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,512,0.14040444956885442
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,65536,128,0.008187555604510838
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,256,0.011773332953453064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,256,0.14319022496541342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,65536,128,0.018223999275101554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,65536,128,0.011695110963450538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,128,0.13249954912397596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,65536,32,0.006157333238257303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,65536,64,0.0181697772608863
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,65536,2048,0.05299199952019585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,65536,32,0.020958221620983545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,16384,65536,0.3291546768612332
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,16384,65536,0.2839031219482422
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,65536,1536,0.1446568833457099
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,16384,16384,0.08583733108308579
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,65536,512,0.015787555111779105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,16384,16384,0.07914755741755168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,65536,0.14702133337656656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,16384,12288,0.0657600031958686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,16384,0.04320533408059014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,65536,256,0.019179556104871962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,16384,10240,0.055085334512922496
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,12288,0.03323910964859857
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,65536,64,0.005945777727497949
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,16384,0.14455200566185847
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,16384,8192,0.04680000080002678
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,16384,10240,0.054937776592042714
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,10240,0.029164443413416546
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,16384,8192,0.045347554816140064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,16384,7168,0.040447112586763166
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,10240,0.14665599664052328
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,8192,0.024792000651359558
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,8192,0.1421111159854465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,16384,7168,0.040361778603659734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,16384,6144,0.03550133440229628
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,7168,0.02219111058447096
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,16384,5120,0.031042665243148804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,7168,0.1422719955444336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,16384,6144,0.036284443404939436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,6144,0.019997333486874897
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,6144,0.14544710848066542
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,65536,0.30952355596754283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,16384,5120,0.03220444586541917
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,16384,4096,0.026499556170569524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,5120,0.017827555537223816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,5120,0.14310133457183838
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,16384,3584,0.02271644439962175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,4096,0.015503111812803479
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,4096,0.1459840006298489
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,16384,12288,0.06261777877807617
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,16384,3072,0.020594666401545208
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,16384,3584,0.026149332523345947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,3584,0.14486222796969944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,3584,0.015134221977657743
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,16384,2560,0.017260443833139207
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,16384,3072,0.023307555251651343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,3072,0.013040000365840064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,12288,0.14123555024464926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,3072,0.1475813388824463
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,16384,2048,0.014465777410401238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,16384,2560,0.021798221601380244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,2560,0.14231110943688288
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,2560,0.01218488895230823
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,16384,1536,0.011719110939237805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,16384,2048,0.0199306673473782
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,2048,0.011013333168294696
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,2048,0.1474471092224121
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,16384,1024,0.00906044410334693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,16384,1536,0.01807822287082672
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,1536,0.009710222482681274
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,16384,768,0.007828444242477417
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,1536,0.1399137841330634
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,16384,1024,0.016419554750124615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,16384,512,0.006210666563775804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,1024,0.14732000562879774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,16384,768,0.014848000473446317
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,768,0.007592000067234039
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,768,0.14293421639336482
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,16384,512,0.013795555465751223
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,512,0.007425777614116669
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,512,0.14619378248850504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,16384,128,0.0038568890757030914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,16384,256,0.013120888835854001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,256,0.00686488880051507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,256,0.14397954940795898
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,16384,64,0.0035342222286595237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,16384,128,0.012768000364303589
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,16384,128,0.13316888279385036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,16384,32,0.003863999827040566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,128,0.0069724445541699724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,16384,64,0.013415999710559845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,16384,32,0.014464888307783337
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,12288,65536,0.24786223305596244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,12288,65536,0.2316044436560737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,12288,16384,0.06465422444873385
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,12288,16384,0.06671644581688775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,65536,0.13849688900841609
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,16384,1024,0.008329778081840938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,12288,12288,0.05036355720625984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,16384,0.03943377733230591
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,16384,256,0.004232888834344016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,12288,12288,0.052821334865358144
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,12288,10240,0.04237066706021627
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,12288,0.0313022236029307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,12288,0.14180089367760554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,12288,10240,0.04650577902793884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,12288,8192,0.03564000129699707
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,16384,4096,0.028134223487642076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,10240,0.027143999934196472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,10240,0.13979910479651556
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,12288,8192,0.03883288966284858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,12288,7168,0.03090933296415541
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,8192,0.0226239992512597
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,12288,6144,0.027610666222042505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,8192,0.1411457724041409
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,65536,0.27505866686503094
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,12288,7168,0.03443822264671326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,7168,0.0207315550910102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,12288,5120,0.023961777488390606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,12288,6144,0.03126488791571723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,6144,0.14166577657063803
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,6144,0.01891911029815674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,12288,5120,0.02788355615403917
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,5120,0.01697688963678148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,5120,0.1396720012029012
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,12288,3584,0.017772444420390658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,12288,4096,0.02383288906680213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,4096,0.014724445011880664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,4096,0.14285155137379965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,12288,3584,0.022282666630215112
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,3584,0.013902222116788229
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,3584,0.1459208859337701
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,16384,0.14676711294386122
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,12288,3072,0.02092622220516205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,3072,0.1406222184499105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,12288,2560,0.013427555561065674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,3072,0.012816888590653738
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,12288,2048,0.011682666838169098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,12288,2560,0.01959822244114346
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,2560,0.011686222420798408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,2560,0.14334755473666722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,12288,1536,0.009964444571071202
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,12288,2048,0.018407111366589863
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,2048,0.1470035579469469
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,2048,0.010403555300500657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,12288,1024,0.008377778033415476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,12288,1536,0.017115554875797696
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,1536,0.009378666679064432
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,1536,0.14093866613176134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,12288,768,0.0069217777086628815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,12288,1024,0.014664888381958008
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,7168,0.14034844769371882
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,1024,0.008016889293988546
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,1024,0.14237956206003824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,12288,512,0.005241777747869492
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,12288,768,0.014192889134089151
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,12288,4096,0.020854221449957952
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,768,0.14078844918145075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,768,0.007382222347789341
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,12288,256,0.0041946665280395085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,12288,512,0.013420444395807056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,512,0.0069573331210348345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,12288,128,0.0038328886859946777
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,512,0.1448248889711168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,12288,256,0.013046222428480784
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,256,0.006638222270541721
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,256,0.1399351093504164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,12288,64,0.0036684444381131064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,12288,3072,0.015826667348543804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,12288,128,0.012720888687504662
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,12288,128,0.13504532972971597
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,12288,64,0.012721777790122561
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,12288,32,0.013401777380042605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,10240,65536,0.2043093310462104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,10240,16384,0.05406755540106031
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,10240,65536,0.22187377346886528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,10240,16384,0.06422399812274508
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,65536,0.1269191106160482
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,16384,0.03853955533769395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,10240,12288,0.042391998900307544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,16384,0.14012711577945286
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,10240,10240,0.03580266568395827
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,10240,12288,0.05123733480771383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,12288,0.029876444074842665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,12288,0.1424257755279541
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,10240,10240,0.0453342232439253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,10240,8192,0.03067555692460802
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,10240,0.025506666964954797
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,10240,0.14198488659328884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,10240,8192,0.0373982224199507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,10240,7168,0.02646133303642273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,12288,32,0.0038275557259718576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,65536,0.2649342219034831
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,8192,0.14225599500868055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,10240,6144,0.02347466680738661
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,10240,7168,0.03357244531313578
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,7168,0.01959111127588484
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,7168,0.1412595510482788
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,10240,6144,0.0304115547074212
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,6144,0.017940445078743827
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,6144,0.14222844441731772
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,10240,5120,0.027071999178992376
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,5120,0.1421911054187351
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,5120,0.016181333197487723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,10240,4096,0.023366222778956097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,10240,3584,0.01481244464715322
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,4096,0.14221066898769802
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,4096,0.014312000738249885
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,10240,3584,0.02207999924818675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,3584,0.013331555657916598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,3584,0.14251466592152914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,10240,3072,0.013242666920026144
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,10240,3072,0.020637333393096924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,3072,0.01237777786122428
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,3072,0.14315466086069742
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,10240,2560,0.011371555427710215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,12288,128,0.006904888898134232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,10240,2048,0.010112889111042023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,8192,0.021451556020312842
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,2560,0.011048888994587792
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,2560,0.1423191097047594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,10240,1536,0.008467555873923832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,10240,2048,0.018001778258217704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,2048,0.14447466532389322
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,10240,1024,0.0069368887278768755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,10240,1536,0.01649777756796943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,1536,0.14025066958533391
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,1536,0.00866844422287411
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,10240,768,0.005984888722499211
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,10240,4096,0.017261332935757108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,10240,1024,0.014472888575659858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,1024,0.14220178127288818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,1024,0.00793511089351442
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,10240,512,0.004282666577233208
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,10240,768,0.013775111072593264
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,768,0.007337777978844113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,768,0.1439324484931098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,10240,256,0.003914666672547658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,10240,512,0.013094222380055321
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,512,0.14312177234225804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,10240,256,0.01276177747382058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,10240,128,0.003543111185232798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,256,0.006618666566080517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,10240,128,0.012418666647540199
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,128,0.13370577494303384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,10240,64,0.003479111111826367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,128,0.006618666566080517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,10240,2560,0.019458666443824768
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,10240,32,0.003549333247873518
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,10240,64,0.012716444830099741
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,2048,0.009667555491129557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,10240,32,0.013079110946920184
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,8192,65536,0.18428356117672393
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,8192,16384,0.04536622100406223
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,8192,65536,0.1777262157864041
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,65536,0.07922044727537367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,8192,16384,0.05391110976537069
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,10240,512,0.007082666787836287
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,8192,12288,0.035784890254338585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,16384,0.025269332859251235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,16384,0.1425457795461019
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,10240,256,0.14494489298926458
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,8192,12288,0.04346755478117201
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,12288,0.019481778144836426
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,8192,10240,0.036927110619015165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,12288,0.14272088474697536
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,8192,8192,0.026103110777007207
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,10240,0.017161778277821012
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,65536,0.24885421329074434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,10240,0.14418755637274847
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,8192,8192,0.031189332405726116
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,8192,7168,0.022258667482270136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,8192,0.014975999792416891
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,8192,0.14789422353108725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,8192,7168,0.028241776757770117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,8192,6144,0.019650666250122916
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,10240,5120,0.020231111182106864
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,7168,0.013930666777822705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,8192,5120,0.017078222499953378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,7168,0.14245333936479357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,8192,6144,0.02516533268822564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,6144,0.013057777451144325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,8192,4096,0.01442311041884952
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,6144,0.14594133694966635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,5120,0.012085333466529846
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,5120,0.13892800278133816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,8192,3584,0.012598222328556908
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,4096,0.010519111322032081
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,8192,4096,0.020909332566791113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,4096,0.147852447297838
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,8192,3072,0.011740444435013665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,8192,3584,0.01958311100800832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,3584,0.14128977722591823
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,3584,0.009993777506881291
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,8192,2560,0.00998488896422916
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,8192,3072,0.018611555298169453
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,3072,0.009011555876996782
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,8192,10240,0.029599110285441082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,3072,0.14273422294192845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,8192,2048,0.008985777695973715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,8192,2560,0.017551110850440133
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,2560,0.008325333396593729
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,8192,1536,0.007873777714040544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,2560,0.14060266812642416
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,2048,0.00796622203456031
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,8192,2048,0.016751110553741455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,8192,1024,0.006216889040337668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,8192,1536,0.014518222875065274
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,1536,0.007304888632562425
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,1536,0.1446453332901001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,8192,768,0.005205333232879639
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,8192,1024,0.013447110851605734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,1024,0.006616888774765863
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,1024,0.13952533404032388
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,8192,512,0.004214222232500712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,8192,768,0.013054221868515015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,768,0.006618666566080517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,8192,256,0.00387999994887246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,8192,512,0.012778666284349231
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,512,0.006234666539563074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,512,0.14919022719065347
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,8192,128,0.003580444388919406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,8192,5120,0.022641777992248535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,8192,256,0.012432888978057437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,256,0.005938666562239329
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,8192,64,0.0034764446318149567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,8192,128,0.011729777687125735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,8192,32,0.003645333151022593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,128,0.1388568878173828
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,8192,128,0.005967111223273807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,8192,64,0.01239466667175293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,8192,32,0.011744889120260874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,7168,65536,0.16396266884273952
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,7168,16384,0.04492977923817105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,7168,65536,0.17483111222585043
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,65536,0.07658577627605863
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,7168,16384,0.05235644512706333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,2048,0.1411857737435235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,7168,12288,0.03587111168437534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,16384,0.1462177832921346
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,7168,12288,0.0421377784676022
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,7168,10240,0.029812445243199665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,12288,0.14352711041768393
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,768,0.14138489299350315
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,65536,0.24713601006401908
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,7168,10240,0.0370053317811754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,7168,8192,0.02624622152911292
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,10240,0.016912000046836007
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,8192,256,0.14050489001803929
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,7168,7168,0.022286223040686712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,8192,0.014800000521871777
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,8192,0.14420089456770155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,7168,7168,0.027655111418830022
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,7168,0.013537777794731988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,7168,0.14136800501081678
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,7168,6144,0.019533332851197984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,6144,0.012778666284349231
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,6144,0.1415839989980062
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,7168,5120,0.017072889539930556
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,16384,0.023530667026837666
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,7168,5120,0.022543110781245764
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,5120,0.011419555379284753
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,5120,0.1398577822579278
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,7168,4096,0.014511111709806653
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,12288,0.018886221779717337
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,7168,3584,0.012063999970753988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,7168,4096,0.020460445019933913
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,4096,0.00999466660949919
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,4096,0.13995110988616943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,7168,3072,0.010928000013033548
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,7168,3584,0.01928622192806668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,3584,0.009114666945404476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,3584,0.14543022049797907
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,7168,8192,0.03050044510099623
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,3072,0.14746756023830837
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,3072,0.008330666356616551
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,7168,2560,0.017486222916179232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,2560,0.00831911133395301
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,7168,2048,0.008943110704421997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,2560,0.14663110838996038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,7168,2048,0.01574222246805827
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,2048,0.007742221984598372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,2048,0.14409866597917345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,7168,6144,0.024978665841950312
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,7168,1024,0.005564444594913059
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,7168,1536,0.01460088955031501
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,1536,0.007258666886223688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,1536,0.14738399452633327
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,7168,768,0.004960888789759742
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,7168,1024,0.013432000246312885
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,1024,0.006610666712125142
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,1024,0.14400355021158853
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,7168,512,0.0041991112132867174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,7168,768,0.012940444052219391
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,768,0.006265777680608962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,768,0.1448648903104994
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,7168,256,0.0038817777401871155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,7168,512,0.012700444294346703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,512,0.006215999937719769
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,512,0.1466417842441135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,10240,0.14184710714552137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,7168,128,0.003567111161020067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,7168,256,0.012415111064910889
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,7168,3072,0.018565333551830716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,256,0.14287643962436253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,256,0.005932444499598608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,7168,64,0.003540444291300244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,7168,32,0.003526222374704149
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,7168,128,0.012068444656001197
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,7168,128,0.005937777873542573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,7168,128,0.13752888308631048
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,7168,64,0.012439111040698158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,7168,32,0.012433778080675336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,6144,16384,0.036644445525275335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,6144,65536,0.1455048852496677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,6144,65536,0.16883733537462023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,6144,16384,0.05092089043723213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,7168,1536,0.0074035558435651995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,65536,0.07404622104432848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,6144,12288,0.029149333635965984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,16384,0.022690667046440974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,6144,10240,0.02436711059676276
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,16384,0.14355644914839003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,6144,12288,0.04043466515011258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,12288,0.0184542222155465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,6144,8192,0.02104088829623328
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,12288,0.1404204501046075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,6144,10240,0.03480266531308492
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,10240,0.016132444143295288
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,6144,7168,0.018175111876593698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,10240,0.14152533478207058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,6144,8192,0.0292577776643965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,8192,0.014121777481502958
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,6144,6144,0.015980444020695157
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,65536,0.24371910095214844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,6144,7168,0.02641955514748891
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,7168,0.14271733495924208
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,6144,5120,0.014146667387750415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,7168,0.01311911145846049
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,6144,6144,0.02436622149414486
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,6144,0.012390221986505719
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,6144,4096,0.012320000264379712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,7168,2560,0.009292444421185387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,5120,0.011104889214038849
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,5120,0.14317422442966038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,6144,3584,0.011165333290894827
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,6144,4096,0.0203146669599745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,4096,0.009649777577983009
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,4096,0.14934578206804064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,6144,3072,0.009962666365835402
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,3584,0.008665777742862701
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,3584,0.1466346714231703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,6144,3072,0.01834755473666721
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,6144,2560,0.00871733327706655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,3072,0.008609777523411645
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,6144,2560,0.016910221841600206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,2560,0.14510311020745173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,6144,2048,0.007611555357774098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,2560,0.00795111142926746
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,8192,0.1410177813635932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,6144,2048,0.01588622232278188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,6144,1536,0.006770666688680649
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,2048,0.14612444241841635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,2048,0.0075911109646161394
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,6144,0.14680888917711046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,6144,1024,0.004931555440028508
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,6144,1536,0.014458666245142618
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,1536,0.007277333074145847
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,1536,0.14654400613572863
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,6144,768,0.00453333349691497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,6144,1024,0.013435555828942193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,1024,0.006629333313968446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,1024,0.1462844477759467
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,6144,512,0.004096888833575779
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,6144,3584,0.019263111882739596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,6144,768,0.012845333251688214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,768,0.006505777852402792
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,768,0.14771821763780382
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,6144,256,0.00350222239891688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,512,0.006166222194830577
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,6144,512,0.01279022213485506
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,512,0.14419466919369167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,6144,128,0.0031804444475306403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,3072,0.13978044191996256
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,256,0.1394373310936822
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,256,0.0059004442559348206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,6144,64,0.003175111073586676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,6144,128,0.01220355596807268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,6144,32,0.003589333345492681
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,6144,128,0.005904888941182031
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,6144,64,0.012107555237081317
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,6144,32,0.012511110968059964
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,5120,65536,0.11948800086975098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,5120,16384,0.03944177760018243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,5120,65536,0.16417955027686224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,65536,0.07175199853049384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,6144,5120,0.02221511138810052
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,16384,0.14078311125437418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,5120,12288,0.03117777903874715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,16384,0.021939555803934734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,5120,12288,0.038740442858801946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,65536,0.24216622776455352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,12288,0.017670222454600863
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,12288,0.1429120037290785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,5120,10240,0.02232355541653103
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,5120,8192,0.019056888090239633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,10240,0.015798222687509324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,5120,10240,0.03354488809903463
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,6144,256,0.012083555261294046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,5120,7168,0.016938666502634685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,5120,8192,0.02836977773242527
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,8192,0.013791110780504016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,6144,128,0.13524088594648573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,8192,0.14312533537546793
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,5120,6144,0.014615111880832247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,5120,7168,0.025976889663272437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,7168,0.14620710743798151
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,5120,5120,0.012833777401182385
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,5120,6144,0.024129778146743774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,6144,0.011741333537631564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,6144,0.14368977811601427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,5120,5120,0.021856889128684998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,5120,0.14137866761949328
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,5120,4096,0.020269332660569083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,5120,3584,0.009957333405812582
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,4096,0.009317333499590555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,5120,3584,0.018960888187090557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,3584,0.14060621791415745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,5120,3072,0.008968888885445064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,3584,0.008632889224423302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,5120,3072,0.018028444714016385
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,10240,0.14414310455322266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,3072,0.14225778314802381
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,5120,2560,0.008413333031866286
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,5120,2560,0.01613066593805949
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,7168,0.012730666332774691
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,2560,0.14106755786471897
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,5120,2048,0.007407111426194509
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,2560,0.007921777665615082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,5120,4096,0.011448889142937131
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,5120,2048,0.0151591118839052
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,5120,0.01092266705301073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,5120,1536,0.006451555424266391
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,2048,0.13894755310482448
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,2048,0.007335111498832703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,4096,0.147725330458747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,5120,16384,0.04996444450484382
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,5120,1536,0.014147554834683737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,1536,0.1438328954908583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,1536,0.006963555183675554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,5120,768,0.00452977791428566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,5120,1024,0.013408000270525614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,1024,0.006609777609507243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,1024,0.13927733898162842
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,5120,512,0.004023111114899318
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,5120,768,0.01292799992693795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,768,0.142758223745558
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,768,0.006308444258239534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,5120,256,0.0035377778112888336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,5120,512,0.012426666915416718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,512,0.005934222290913264
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,512,0.14452889230516222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,5120,128,0.0032097777972618737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,5120,256,0.012090666426552666
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,256,0.005920889063013925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,256,0.14184178246392146
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,5120,64,0.003167111012670729
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,3072,0.008015111088752747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,5120,128,0.012079999678664737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,5120,128,0.005954666684071223
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,5120,128,0.13159644603729248
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,5120,64,0.01237777786122428
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,4096,65536,0.0958097775777181
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,4096,16384,0.0322035551071167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,65536,0.06958578030268352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,4096,16384,0.038387556870778404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,5120,1024,0.004872888740566042
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,4096,12288,0.02574311031235589
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,16384,0.020631111330456205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,4096,12288,0.030650665362675984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,12288,0.14593422412872314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,4096,10240,0.01921688848071628
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,12288,0.017154667112562392
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,4096,10240,0.02701777716477712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,5120,32,0.0031991110493739447
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,10240,0.14684000280168322
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,4096,8192,0.016743999388482835
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,10240,0.015371556083361307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,5120,32,0.012104888757069906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,4096,8192,0.023407111565272015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,4096,7168,0.014479999740918478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,8192,0.13984978199005127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,4096,7168,0.021327111456129286
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,7168,0.012471111284361945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,7168,0.1421360042360094
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,4096,6144,0.020024889045291476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,16384,0.14192710982428655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,6144,0.14107822047339544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,4096,5120,0.011388444238238864
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,6144,0.011342222491900126
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,4096,5120,0.01853155593077342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,4096,4096,0.01071999967098236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,5120,0.14334844218360052
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,5120,0.010103999740547603
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,4096,4096,0.0169813334941864
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,4096,0.008734222087595198
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,4096,0.1416106621424357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,4096,3584,0.0158915552828047
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,3584,0.14277066124810114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,4096,3072,0.008595555192894405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,3584,0.00833511104186376
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,8192,0.013386666774749756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,4096,3072,0.015139554937680563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,4096,2560,0.007619555625650618
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,3072,0.14364977677663168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,3072,0.008249777886602614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,4096,2048,0.006545777950021956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,4096,6144,0.013032889200581444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,4096,2560,0.01442488862408532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,2560,0.0076044441925154785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,2560,0.1453146669599745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,4096,1536,0.0052373334765434265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,4096,2048,0.013599110974205865
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,65536,0.24063467979431152
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,2048,0.14712266127268472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,4096,1024,0.004193777839342753
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,4096,1536,0.01311288856797748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,1536,0.143157336446974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,1536,0.006950221955776215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,4096,768,0.0041546668443414895
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,4096,1024,0.01274222218328052
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,1024,0.006616000086069107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,1024,0.14032710923088923
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,4096,512,0.0036880001425743103
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,4096,768,0.012436444560686747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,768,0.006262222097979651
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,768,0.14290577835506862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,4096,256,0.003156444471743372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,4096,512,0.01238222254647149
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,512,0.0059279998143513995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,512,0.14386667145623103
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,4096,128,0.0031884445084465873
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,256,0.14209422800276014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,256,0.005912888795137405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,4096,128,0.011752000285519494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,4096,64,0.003192000091075897
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,128,0.005945777727497949
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,4096,128,0.1345857779184977
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,4096,32,0.003212444484233856
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,4096,65536,0.12222399976518418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,4096,64,0.012067555553383298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3584,65536,0.08570666445626153
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,4096,32,0.01202666676706738
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3584,16384,0.028939555088678997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3584,65536,0.12112443976932102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,4096,2048,0.007248000138335758
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,65536,0.06867377625571357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3584,16384,0.03604799840185378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3584,12288,0.02440266642305586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,16384,0.1481706698735555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,16384,0.020506666766272653
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3584,10240,0.02128177715672387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3584,12288,0.029157333903842505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,12288,0.1398222181532118
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,65536,0.2394568920135498
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,12288,0.016912889149453905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3584,8192,0.015107555521859063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3584,10240,0.02659911082850562
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,10240,0.14566311571333143
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,4096,3584,0.009188444250159794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3584,7168,0.013491555220550962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,4096,256,0.01202933324707879
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,8192,0.012874667015340594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,8192,0.1448568900426229
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3584,6144,0.01236533373594284
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,7168,0.145870221985711
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3584,6144,0.01980622278319465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,6144,0.011125333607196808
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3584,5120,0.018423111902342904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,5120,0.14513866106669107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,5120,0.010403555300500657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3584,4096,0.010197333163685268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3584,4096,0.016767111089494493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,10240,0.015097777048746744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3584,8192,0.022925333844290838
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3584,3584,0.008984000318580205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3584,7168,0.020983111527231004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,4096,0.008662222160233391
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,4096,0.1451359987258911
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,7168,0.011619555453459421
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3584,3072,0.008098666866620382
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3584,3584,0.015573332707087198
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,3584,0.00831733312871721
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,3584,0.1416728893915812
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3584,2560,0.007249777515729268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3584,3072,0.01513155632548862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,3072,0.14441066318088108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,3072,0.008088889221350351
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3584,2048,0.006606222026877933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3584,2560,0.01443288889196184
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,2560,0.007605333295133378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,2560,0.14028088251749674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3584,1536,0.005550222264395819
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3584,2048,0.013456000222100152
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,2048,0.14544710848066542
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,2048,0.007312888900438945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3584,1024,0.0041831110914548235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3584,1536,0.01310755560795466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,1536,0.0069546666410234236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,1536,0.14335555500454372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3584,768,0.0038684445122877755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3584,1024,0.01258044441541036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,1024,0.006621333460013072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,1024,0.14604088995191786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3584,512,0.0035422220826148987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3584,768,0.012427555190192329
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,768,0.14928800529903835
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,768,0.006239999913507038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3584,256,0.0034888887570963967
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3584,512,0.01205599970287747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,512,0.005933333188295364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,512,0.14115644825829401
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3584,128,0.00349866681628757
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3584,256,0.011737777955002256
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,256,0.14178133010864258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3584,64,0.0031991110493739447
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3584,32,0.0031759999692440033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3584,128,0.011699555648697747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,128,0.0059573331640826324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,128,0.13160267141130236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3584,64,0.011757333245542316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3072,65536,0.07639022005928887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3072,16384,0.025342222717073228
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3584,5120,0.01109777804878023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3072,65536,0.11842666731940375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,65536,0.06749422020382352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3072,16384,0.034733331865734525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,16384,0.02017155620786879
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,16384,0.145651552412245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,65536,0.2373911009894477
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3072,12288,0.02903111113442315
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,12288,0.14816266960567898
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3072,10240,0.017868444323539734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,12288,0.01647822227742937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3072,10240,0.02604355580276913
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,10240,0.014863999353514777
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3072,8192,0.016018667154841952
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3072,8192,0.02242044442229801
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,8192,0.14484800232781306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3072,7168,0.01311288856797748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,8192,0.012479999827014076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3584,32,0.011056889262464313
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3072,7168,0.020975111259354487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,7168,0.1397893296347724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3072,6144,0.011714666253990598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,7168,0.011723555624485016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3072,6144,0.019712888532214694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3072,12288,0.021193777521451313
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,6144,0.14499555693732366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3072,5120,0.01035377797153261
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3072,5120,0.017614222235149808
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,5120,0.009683555199040307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,5120,0.14605244000752768
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3072,4096,0.009422221945391761
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3072,4096,0.015832000308566622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,4096,0.00869599978129069
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3072,3584,0.008625778059164682
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,4096,0.14401777585347494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,10240,0.14051822821299234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3072,3072,0.00813600007030699
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3072,3584,0.014907555447684394
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,3584,0.008354666332403818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,3584,0.1417377789815267
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,3584,256,0.005894222193294101
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3072,2560,0.0069511110583941145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3072,3072,0.014482667048772177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,3072,0.1425857808854845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3072,2048,0.006020444548792309
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3072,2560,0.014091556270917257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,2560,0.14111199643876818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,2560,0.007784888976150089
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3072,1536,0.00519822206762102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3072,2048,0.013419555293189155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,3584,6144,0.1419644488228692
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,2048,0.14143644438849554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3072,1536,0.013063111239009433
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3072,1024,0.004509333521127701
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,1536,0.006672888994216919
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,1536,0.13959733645121256
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,6144,0.010383110907342698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3072,768,0.004173333446184794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3072,1024,0.012717333104875354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,1024,0.14238399929470485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,1024,0.006329777754015393
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3072,512,0.003843555433882607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3072,768,0.012082666986518435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,768,0.006384000182151794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,768,0.1414088937971327
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3072,256,0.003527111063400904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3072,512,0.01240533341964086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,512,0.141120884153578
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,512,0.00627377794848548
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3072,128,0.003149333306484752
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3072,256,0.011717333561844297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,256,0.00589155571328269
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3072,64,0.0031742221779293488
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3072,128,0.011403555671374003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,128,0.0058959999846087555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,3072,32,0.0031786666562159858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,128,0.13333777586619058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3072,64,0.011764444410800934
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2560,65536,0.06426133049858941
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,3072,0.007982222570313348
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,3072,32,0.01146666705608368
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2560,16384,0.022964444425370958
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2560,65536,0.11554666360219319
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2560,16384,0.03389066788885329
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,65536,0.06609333223766752
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2560,12288,0.019488889310095046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,3072,2048,0.007298666569921706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,16384,0.14461599455939397
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2560,10240,0.01720355616675483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,12288,0.015941333439615037
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2560,12288,0.028074665202034846
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,12288,0.14229867193433973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2560,10240,0.025756445195939805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,10240,0.14378400643666586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,10240,0.014461333552996317
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2560,7168,0.013975999421543546
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,8192,0.012068444656001197
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2560,7168,0.020282665888468426
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,3072,256,0.14127822717030844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2560,6144,0.013073777986897362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2560,6144,0.01924800044960446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,6144,0.14007555113898382
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,16384,0.020078221956888836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,6144,0.010323555933104621
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,65536,0.2384186585744222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2560,8192,0.015150222513410779
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2560,5120,0.01794577803876665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,5120,0.14260889424218073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2560,8192,0.021958221991856892
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2560,4096,0.009326222042242685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,8192,0.14653866820865208
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,5120,0.009678222239017487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2560,3584,0.008412444757090675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2560,4096,0.015830222103330824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,4096,0.008632889224423302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,7168,0.14386933379703098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2560,3584,0.01514311134815216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2560,3072,0.007311111523045435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,7168,0.011049778097205691
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,3584,0.008619555168681675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2560,2560,0.006799110935793982
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2560,3072,0.014455111490355598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,3072,0.007932444413503012
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,3072,0.14338933096991643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2560,2048,0.005702222386995952
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2560,2560,0.013796444568369122
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,2560,0.007582222421964009
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,2560,0.14104621940188938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2560,1536,0.005196444276306364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2560,2048,0.01348888874053955
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,2048,0.007256888680987888
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,2048,0.14490044116973877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2560,1024,0.0041884444653987885
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2560,1536,0.013071111506885953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2560,5120,0.010388444695207808
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,1536,0.0069475554757648045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,1536,0.14418221844567194
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2560,768,0.003882666842805015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2560,1024,0.012493333054913415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,1024,0.006578666468461354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,1024,0.1397155523300171
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2560,512,0.003536000019974179
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2560,768,0.01240888900227017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,768,0.13923733764224583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2560,512,0.012087999946541257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2560,256,0.003463111196955045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,512,0.005980444451173146
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,512,0.14213244120279947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2560,128,0.003169777699642711
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2560,256,0.01185066666867998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,256,0.005935110979610019
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,256,0.1384613381491767
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2560,64,0.003156444471743372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,3584,0.14269777139027914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,128,0.005923555543025334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,128,0.13314667012956408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2560,32,0.003149333306484752
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2560,64,0.011677333050303988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2048,65536,0.05325422353214688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2560,32,0.011732444167137146
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2048,16384,0.019261333677503798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2048,65536,0.11198577615949844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,65536,0.06451821989483304
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2048,16384,0.033253331979115806
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,16384,0.1412231127421061
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2048,12288,0.015488889482286243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,65536,0.23761333359612358
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,16384,0.019574221637513902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2048,10240,0.01455200049612257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2048,12288,0.027995556592941284
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,12288,0.015264888604482016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2048,10240,0.02555822167131636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,2560,768,0.006246222390068903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2048,8192,0.013841778039932251
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,10240,0.01310488912794325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,10240,0.1428568893008762
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2048,7168,0.012498666842778524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2048,8192,0.021700445148679946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,2560,4096,0.14356266127692327
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,8192,0.01145244472556644
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,8192,0.13996622297498915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2048,6144,0.011436444189813403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2048,7168,0.01961511042382982
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2560,128,0.011745777395036487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,7168,0.011065777805116443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2048,6144,0.0188711103465822
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2048,5120,0.010303110712104374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,6144,0.13899200492435032
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,6144,0.01035199976629681
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2048,4096,0.008638222184446123
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2048,5120,0.01722311145729489
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,5120,0.009659555223253038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2048,3584,0.008225777910815345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2048,4096,0.015152888165579902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,4096,0.141393780708313
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2048,3072,0.0069066666894488865
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2048,3584,0.015086222026083203
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,3584,0.14240177472432455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,3584,0.00833155545923445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2048,2560,0.006272000157170826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2048,3072,0.014338667194048563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,3072,0.007948444121413762
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,3072,0.14201422532399496
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2048,2048,0.005239999956554837
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2048,2560,0.013784888717863293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,2560,0.007465778125656976
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,12288,0.14570666684044734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,2560,0.14067110750410292
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2048,1536,0.004571555389298333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2048,2048,0.013394667042626275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,2048,0.007313778003056844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2048,1536,0.013096888860066732
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2048,1024,0.004231111043029361
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,1536,0.14241154988606772
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,1536,0.006950221955776215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2048,1024,0.012404444317022959
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,1024,0.0064560001095136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,7168,0.1455635494656033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,1024,0.14262133174472386
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2048,768,0.0038568890757030914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2048,512,0.003552888830502828
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,768,0.006237333433495627
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,5120,0.1399351093504164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,768,0.1483057737350464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2048,512,0.011768889096048145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2048,256,0.0032133333798911837
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,512,0.005920889063013925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,512,0.14325422710842556
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2048,256,0.011715555356608497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2048,128,0.003125333330697484
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,256,0.1461377806133694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,256,0.00591466658645206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2048,64,0.0028648889727062653
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2048,128,0.011687111523416309
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,128,0.13557867209116617
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,128,0.005904888941182031
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2048,64,0.011727111207114326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1536,65536,0.04360711243417528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2048,32,0.011706666813956367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,2048,2048,0.15026844872368708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1536,16384,0.014635556273990206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,65536,0.0633902218606737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,65536,0.23681243260701498
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1536,16384,0.032610668076409235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1536,12288,0.014104889498816596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,16384,0.1431866619322035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,16384,0.018903111418088276
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,2048,768,0.012415111064910889
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1536,12288,0.027482666903071936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1536,10240,0.012462221913867526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,12288,0.14158222410413954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,2048,4096,0.00867022242810991
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1536,8192,0.011232000258233813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1536,10240,0.024846222665574815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,10240,0.013247999880048962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1536,8192,0.021232888102531433
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,8192,0.146978669696384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1536,7168,0.010419555836253695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,2048,32,0.0028159999185138275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1536,7168,0.01940355532699161
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,7168,0.13966133859422472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,7168,0.010680889089902243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1536,6144,0.010402667025725046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1536,65536,0.11009599765141805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,6144,0.010036444498433007
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,6144,0.14290222856733534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1536,5120,0.009512888888518015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1536,5120,0.01651022169325087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,5120,0.009367110828558603
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1536,4096,0.008662222160233391
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,5120,0.1409991052415636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1536,4096,0.01514933341079288
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,4096,0.008587555752860175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1536,3584,0.007288888924651676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,4096,0.14598133828904894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,12288,0.014428445034556918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1536,3072,0.0069102222720781965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,3584,0.00830844458606508
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,3584,0.1395297712749905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,10240,0.1462471087773641
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1536,2560,0.00611200018061532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1536,3072,0.014215999179416232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,3072,0.0079297779334916
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,8192,0.011408000356621213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1536,2048,0.0051893335249688895
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1536,2560,0.013831111292044321
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,2560,0.14581688245137533
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1536,1536,0.004927999857399198
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1536,2048,0.013414222333166333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,2048,0.007279111279381647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1536,1536,0.01293333371480306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1536,6144,0.01817244456874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,1536,0.1443528864118788
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1536,1024,0.004202666795916027
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1536,1024,0.01238666640387641
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,1024,0.14259200625949436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1536,768,0.012411555482281579
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1536,3584,0.014767110347747803
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,768,0.14058755503760442
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1536,512,0.0034924443397257063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,768,0.006249777972698212
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,3072,0.1455920007493761
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1536,512,0.012075555821259817
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1536,256,0.003176888864901331
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,2560,0.007600000335110559
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,512,0.005969777703285217
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,2048,0.14443555143144396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1536,256,0.012042666474978128
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1536,128,0.0028702221396896574
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,256,0.13922399944729275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,256,0.005899555567238066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1536,128,0.012048000262843238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1536,64,0.0028080000645584534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,128,0.005626666463083691
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,1024,0.006287999865081575
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1536,32,0.002854222224818336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1536,768,0.0038311113086011675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1024,65536,0.030454221698972914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1536,64,0.01181244436237547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1536,32,0.011706666813956367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1024,16384,0.012982221941153208
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1024,65536,0.10660800006654526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,65536,0.0606879989306132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1024,16384,0.032279110617107816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,65536,0.23905510372585723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,16384,0.14031467172834608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,16384,0.017200888858901132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,12288,0.014121777481502958
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,12288,0.13980444272359213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1024,10240,0.012359999948077731
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,512,0.14096621672312418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1024,10240,0.024396444360415142
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1024,8192,0.011250666446155973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,10240,0.013068444199032254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,1536,1536,0.0069573331210348345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1024,8192,0.020455999506844413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,8192,0.14171732796563044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,1536,128,0.13181066513061523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,8192,0.011404444773991903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1024,7168,0.018868444694413077
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,7168,0.010696888797812991
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1024,6144,0.009360888765917884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1024,6144,0.01792355544037289
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,6144,0.14154222276475695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1024,5120,0.00867555538813273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1024,12288,0.013737777868906656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1024,5120,0.016537777251667447
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1024,4096,0.007646222081449296
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,5120,0.14607644081115723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,5120,0.009663110805882348
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1024,12288,0.02683911058637831
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1024,3584,0.007259555160999298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1024,4096,0.01512977812025282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,4096,0.00833777752187517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,4096,0.14757243792215982
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1024,3072,0.006595555692911148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1024,3584,0.01482399966981676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,10240,0.1433511045244005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,3584,0.008282666405042013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,3584,0.14168266455332437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1024,2560,0.005899555567238066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1024,3072,0.014118222726715935
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,3072,0.00793333351612091
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,3072,0.14269155926174587
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1024,2048,0.005249777601824866
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1024,7168,0.010302222437328762
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,2560,0.007295110987292395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1024,1536,0.004566222015354368
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1024,2048,0.013406222065289816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,2048,0.00723555518521203
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,2048,0.14337066809336343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,7168,0.13971911536322698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1024,1024,0.0038728887836138406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1024,1536,0.012768000364303589
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,1536,0.006911110960774952
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,1536,0.14295289251539442
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1024,768,0.0034942221310403612
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1024,1024,0.012441777520709567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,1024,0.006272888845867581
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,1024,0.14443911446465388
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1024,512,0.003519111209445529
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1024,768,0.01240888900227017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,768,0.1405128902859158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1024,256,0.003168888803985384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1024,512,0.01204711116022534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,512,0.005920889063013925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1024,256,0.011678222152921887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,256,0.1418666707144843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1024,128,0.002830222249031067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1024,128,0.011724444727102915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,128,0.13023200300004747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1024,64,0.0028657778683635923
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,128,0.005892444401979446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,1024,32,0.0028479999552170434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,2560,0.14566844039493138
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1024,64,0.011608888705571493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,768,65536,0.026995556222067937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1024,32,0.011701333026091257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,768,65536,0.10428622033860947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,768,16384,0.013608000344700284
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,65536,0.23601688279045951
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,768,0.006239111224810283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,768,16384,0.03164888752831353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,768,12288,0.012268444730175866
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,16384,0.1447991132736206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,1024,512,0.1473742193645901
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,768,12288,0.026047110557556152
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,12288,0.14486666520436606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,768,10240,0.011019555230935415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,12288,0.014161777165200977
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,768,10240,0.02380888826317257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,10240,0.14008711444007024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,768,8192,0.010034666293197209
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,10240,0.013066666821638743
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,768,8192,0.019884443945354886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,8192,0.14528444078233507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,768,7168,0.008998221821255153
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,8192,0.011405333048767514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,65536,0.05849066707823011
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,768,7168,0.018767111831241183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,768,6144,0.00832088871134652
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,6144,0.010057777994208867
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,768,6144,0.01755822201569875
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,6144,0.14492444197336832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,768,5120,0.00795111142926746
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,1024,256,0.00572622236278322
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,768,5120,0.016542222764756944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,5120,0.14352177249060735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,768,4096,0.007923555870850882
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,5120,0.009526222116417354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,1024,2560,0.013767110804716745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,768,4096,0.015171556009186639
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,768,3584,0.007618666523032718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,4096,0.008472888833946651
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,768,3584,0.014818666709793938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,768,3072,0.006325333482689328
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,3584,0.008015111088752747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,7168,0.010679999987284342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,768,3072,0.01442488862408532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,7168,0.14052089055379233
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,768,2560,0.005933333188295364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,3072,0.14730844232771131
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,3072,0.00794044468137953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,768,2560,0.013786666923099093
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,768,2048,0.005183111048407025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,2560,0.007289778027269576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,6144,0.010061333576838175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,768,1536,0.004572444491916233
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,2048,0.006976889239417181
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,768,1536,0.01273155543539259
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,1536,0.14834400018056235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,768,1024,0.0041724443435668945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,1536,0.006597333484225803
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,4096,0.1451688872443305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,768,1024,0.012422222230169507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,768,768,0.0035235554807715942
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,1024,0.006623111251327727
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,3584,0.14206577671898737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,768,768,0.012069332930776807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,768,512,0.0032035555276605818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,768,0.006245333287451003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,768,0.1449928945965237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,768,512,0.011738667057620155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,512,0.0059262220230367445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,16384,0.01715288890732659
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,512,0.14085866345299616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,2560,0.14243555068969727
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,768,128,0.0028133332315418455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,768,256,0.012079999678664737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,256,0.005882666756709416
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,768,2048,0.013412444127930535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,768,64,0.002502222235004107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,768,128,0.011755555868148804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,768,32,0.002798222212327851
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,768,64,0.011390222443474663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,512,65536,0.020271110865804885
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,768,32,0.011521778172916837
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,512,65536,0.10246400038401286
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,1024,0.1446568833457099
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,512,16384,0.011298666397730509
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,65536,0.23818310101826987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,65536,0.058152887556287974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,512,16384,0.030491554074817236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,512,12288,0.009651555783218807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,16384,0.14269777139027914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,16384,0.017134222719404433
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,768,256,0.0028488888508743713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,512,12288,0.025026667449209426
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,12288,0.014420444766680399
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,512,10240,0.02290311124589708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,256,0.1387795607248942
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,10240,0.14081954956054688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,2048,0.14475644959343806
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,768,128,0.005897777775923411
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,10240,0.012903110848532783
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,768,128,0.13638667265574136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,512,7168,0.008281778130266402
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,512,8192,0.019578667150603402
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,8192,0.142922666337755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,8192,0.011355555719799466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,512,6144,0.007624000310897827
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,512,7168,0.0189057770702574
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,7168,0.01072799993885888
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,7168,0.14619378248850504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,512,5120,0.008143111235565608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,512,6144,0.017516444126764934
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,6144,0.14384178320566812
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,512,5120,0.016188444362746347
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,512,4096,0.007611555357774098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,5120,0.009504000345865885
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,512,4096,0.014948444234000312
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,4096,0.14182399378882513
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,512,3584,0.0069439998931354946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,4096,0.008635555704434713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,512,10240,0.009677333136399588
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,512,3584,0.01479200025399526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,512,3072,0.006593777901596493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,3584,0.14559466309017605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,3584,0.008293333152929941
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,512,8192,0.008672888908121321
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,512,2560,0.005600888695981767
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,512,3072,0.014412444498803882
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,3072,0.007715555528799693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,3072,0.14705067210727268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,512,2048,0.005218666460778978
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,512,2560,0.013769778112570444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,2560,0.1431608862347073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,2560,0.007409777906205919
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,512,1536,0.004541333350870344
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,512,2048,0.013073777986897362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,2048,0.007260444263617198
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,2048,0.1421626673804389
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,6144,0.010083555347389644
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,512,1024,0.0038497779104444715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,512,1536,0.01276533305644989
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,1536,0.14672444926367864
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,1536,0.0068862222962909276
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,512,768,0.0035075553589397003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,512,1024,0.012413333687517377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,1024,0.006447111152940326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,1024,0.1406737830903795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,512,512,0.0031991110493739447
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,512,768,0.01237777786122428
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,768,0.006259555617968242
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,768,0.1459368864695231
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,512,512,0.012054222325483957
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,512,0.14869244893391928
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,512,0.005939555664857228
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,512,128,0.002805333377586471
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,12288,0.14301599396599665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,512,256,0.011436444189813403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,256,0.00563822231358952
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,256,0.14012266529930964
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,512,64,0.0026444445053736367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,512,32,0.002830222249031067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,512,128,0.011741333537631564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,512,128,0.005910222315125995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,128,0.13496355215708414
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,512,64,0.011695110963450538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,512,32,0.0110880004035102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,256,16384,0.009684444301658207
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,256,65536,0.10028177499771118
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,65536,0.052128000391854175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,65536,0.23773066202799478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,256,16384,0.030181331766976252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,256,12288,0.008452444440788692
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,16384,0.017136888371573556
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,16384,0.1492568916744656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,256,10240,0.009015999734401703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,256,12288,0.025051555699772302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,12288,0.1417457792494032
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,512,5120,0.14111910925971136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,256,8192,0.008263111114501953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,12288,0.014076444837782117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,256,10240,0.022639110684394836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,10240,0.0129439996348487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,256,7168,0.007701333198282454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,512,256,0.0031928889867332247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,8192,0.011353777514563667
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,256,8192,0.019552888141738046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,8192,0.1398035552766588
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,256,6144,0.007237333390447829
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,256,7168,0.018876444962289598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,7168,0.010674667027261523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,7168,0.14513066079881456
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,256,6144,0.01752355529202355
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,256,5120,0.007954667011896769
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,6144,0.01019022199842665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,6144,0.14119377401140001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,256,4096,0.007304000357786815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,256,65536,0.016207999653286405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,256,5120,0.016534222496880423
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,5120,0.009348444640636444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,5120,0.1402719948026869
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,256,3584,0.0070364442136552595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,256,4096,0.015034667319721647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,256,3072,0.006606222026877933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,4096,0.00848711116446389
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,256,3584,0.01443199978934394
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,3584,0.008273777862389883
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,3584,0.1401146650314331
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,256,2560,0.006463999963468975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,256,3072,0.013873777455753751
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,3072,0.00796977761718962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,256,2048,0.006267555471923616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,3072,0.14033955997890896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,256,2560,0.013775111072593264
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,2560,0.1402240064409044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,256,1536,0.004468444320890638
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,256,2048,0.013433777623706393
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,2048,0.007248889240953658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,256,1536,0.013003555436929068
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,256,1024,0.003818666769398583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,1536,0.14585422144995794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,10240,0.14647111627790663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,1024,0.006302222195598815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,256,1024,0.012435555458068848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,1024,0.14716088771820068
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,256,768,0.012327999704413943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,768,0.14511822329627141
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,256,512,0.0031600000543726813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,768,0.006220444622966979
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,256,512,0.012069332930776807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,4096,0.14525332715776232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,512,0.14049243927001953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,512,0.005966222120655908
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,256,256,0.011743110915025076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,2560,0.007511999871995714
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,256,128,0.0028115556471877625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,256,0.14651910463968912
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,256,0.005892444401979446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,2048,0.14118311140272352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,256,64,0.0028213332924577924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,128,0.0058853332367208265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,256,128,0.135897782113817
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,256,32,0.0028560000161329904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,128,65536,0.015321777926550971
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,256,64,0.011410666836632622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,256,32,0.011515555282433828
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,256,768,0.003526222374704149
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,128,16384,0.009385777844323052
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,128,65536,0.099108444319831
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,65536,0.05048888921737671
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,65536,0.23652267456054688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,128,12288,0.00829777783817715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,128,16384,0.02981333269013299
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,16384,0.14026667012108698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,16384,0.01701955497264862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,128,10240,0.007753777835104201
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,128,12288,0.025036444266637165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,12288,0.014099554883109199
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,256,256,0.0031475555151700974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,128,10240,0.022992889086405437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,128,8192,0.007253333098358578
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,10240,0.14174666669633654
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,10240,0.01276000009642707
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,128,7168,0.0069324444565508105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,128,8192,0.019655111763212416
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,8192,0.011369778050316704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,8192,0.14215200477176243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,128,7168,0.0188746667570538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,7168,0.01069866700304879
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,256,128,0.011354666617181567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,7168,0.14474933677249485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,128,5120,0.006276444428496891
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,256,1536,0.0069066666894488865
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,128,6144,0.017550221747822232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,6144,0.010009778042634329
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,6144,0.14137778017255995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,128,4096,0.0063457778758472866
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,5120,0.009356444080670675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,5120,0.14179910553826228
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,128,3584,0.006274666637182236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,128,4096,0.014965333872371249
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,4096,0.00832622249921163
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,4096,0.1484960052702162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,128,3072,0.006230222268237009
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,128,3584,0.014815110299322339
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,3584,0.14164800114101833
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,128,3072,0.014096889230940076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,128,2560,0.006552000012662675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,12288,0.1488435533311632
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,3072,0.00794577764140235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,3072,0.1442355579800076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,128,2048,0.005903111149867375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,128,2560,0.01348444405529234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,2560,0.007306666837798224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,2560,0.13861689302656385
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,128,1536,0.004326222257481681
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,128,2048,0.013409777647919126
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,2048,0.1420782274670071
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,128,1024,0.0038577777643998465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,128,1536,0.01272266689274046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,1536,0.006795555353164673
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,1536,0.1476800044377645
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,128,768,0.0035137778355015647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,128,1024,0.012415111064910889
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,1024,0.00656088896923595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,1024,0.13883645004696316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,128,512,0.0031973332580592898
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,128,768,0.011741333537631564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,768,0.006263111200597551
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,768,0.14208088980780706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,128,5120,0.016391999191708036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,128,256,0.0028640000770489373
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,128,512,0.01202044470442666
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,512,0.14771200550927058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,512,0.0059368887709246735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,128,128,0.0028364445186323593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,128,256,0.011729777687125735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,256,0.13991110854678684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,256,0.005629333357016246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,128,64,0.0024951110697454875
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,3584,0.007992000215583378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,128,128,0.011683555940786997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,128,32,0.0026177778426143858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,128,0.005899555567238066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,128,128,128,0.14078933662838405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,64,65536,0.014127110441525778
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,64,16384,0.008101333346631791
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,128,64,0.010791999598344168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,64,12288,0.007641777396202087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,128,32,0.011406222151385413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,64,10240,0.007342221836249034
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,64,8192,0.006650666809744305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,64,16384,0.02976977825164795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,64,12288,0.024694222542974684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,64,10240,0.022679110368092854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,128,128,2048,0.007095110913117726
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,64,8192,0.019663110375404358
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,64,5120,0.006248000181383557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,128,6144,0.006567111031876669
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,64,4096,0.00563822231358952
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,64,6144,0.01722755531469981
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,64,3584,0.006229333165619109
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,64,5120,0.016526222229003906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,64,3072,0.006272888845867581
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,64,4096,0.015159999330838522
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,64,2560,0.006258666515350342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,64,3072,0.014117333624098035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,64,2048,0.006131555471155379
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,64,2560,0.013805333111021252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,64,1536,0.0042204442951414315
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,64,2048,0.013438222308953604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,64,1024,0.0038488888078265717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,64,1536,0.013054221868515015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,64,768,0.0035111109415690103
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,64,1024,0.012447111308574677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,64,512,0.0032293332947625052
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,64,768,0.01237066669596566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,64,256,0.0028568889117903183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,64,7168,0.006637333167923822
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,64,128,0.0028035555862718155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,64,512,0.012105777859687805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,64,6144,0.006597333484225803
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,64,64,0.0028533333291610083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,64,256,0.011747555600272285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,64,7168,0.018215111560291715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,64,32,0.002489777695801523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,64,128,0.011389333340856763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,32,65536,0.013438222308953604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,32,16384,0.007987555530336168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,64,64,0.011351111034552256
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,64,32,0.011417778001891242
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,32,12288,0.007643555601437886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,32,10240,0.007240888973077138
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,32,8192,0.006650666809744305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,32,65536,0.0959048867225647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,32,16384,0.029693331983354356
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,32,7168,0.0069520001610120135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,32,12288,0.02500088844034407
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,32,6144,0.006259555617968242
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,64,3584,0.014141332772043018
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,32,5120,0.006227555374304454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,32,8192,0.019525332583321463
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,32,4096,0.0059075554211934405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,32,3584,0.006212444355090459
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,32,7168,0.018897778458065458
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,32,3072,0.0057706667317284485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,32,5120,0.015863110621770222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,32,2560,0.006267555471923616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,32,4096,0.014779556128713818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,32,2048,0.005929777605666055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,32,3584,0.014589332871966891
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,32,1536,0.0042408886882993905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,32,3072,0.014113777213626437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,32,2560,0.01314133322901196
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,32,2048,0.013071111506885953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,32,1536,0.012772444221708508
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,32,512,0.0031999999450312722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,32,768,0.012055111428101858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,32,256,0.0028168888141711554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,64,65536,0.09672622548209296
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,32,128,0.002833777831660377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,32,256,0.011712888876597086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,32,64,0.002521777732504739
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,32,128,0.011381333072980246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,32,32,0.002826666666401757
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,32,64,0.011408000356621213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,32,32,0.011376889215575324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,32,10240,0.022749332918061152
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,32,6144,0.017551110850440133
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,65536,16384,0.3739039897918701
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,65536,16384,0.21803822782304552
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,32,1024,0.0038577777643998465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,128,32,768,0.0034728890491856467
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,32,1024,0.012455999851226807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,65536,12288,0.28078312344021267
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,65536,12288,0.16190132829878065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,128,32,512,0.011779555843936073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,16384,0.13856355349222818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,65536,10240,0.20733866426679823
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,65536,10240,0.13769511381785074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,12288,0.10188266966078018
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,10240,0.08597422308391994
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,65536,8192,0.11745333671569824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,65536,8192,0.17405333783891466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,65536,7168,0.15105866061316595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,8192,0.07027200195524427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,65536,7168,0.10032800171110366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,12288,0.24661244286431205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,65536,6144,0.12926667266421848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,7168,0.06706844435797797
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,10240,0.21684088971879747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,16384,0.31692179044087726
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,8192,0.18314045005374482
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,65536,6144,0.0884551140997145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,6144,0.058111998769972056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,65536,5120,0.07641777727339003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,7168,0.1684115595287747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,5120,0.046816888782713145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,65536,4096,0.09142222007115682
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,65536,3584,0.07973866992526583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,6144,0.15444355540805393
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,5120,0.14741955863104925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,65536,3584,0.05764000283347236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,65536,3072,0.06958488623301189
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,4096,0.1427324480480618
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,3584,0.14770755502912733
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,65536,3072,0.05111999975310432
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,3072,0.03118844495879279
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,65536,2560,0.058232002788119845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,3072,0.14796445104810926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,65536,2560,0.04583200150065952
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,65536,5120,0.109497778945499
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,65536,2048,0.04843022094832527
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,2560,0.027813331948386297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,65536,1536,0.03743911120626662
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,2560,0.14346578386094835
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,2048,0.023737778266270954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,2048,0.14438045024871826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,65536,1536,0.03359466791152954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,65536,4096,0.0639608899752299
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,65536,1024,0.02795999911096361
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,1536,0.1424702273474799
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,4096,0.039136889908048846
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,65536,1024,0.028586665789286297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,1024,0.14401777585347494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,65536,768,0.021240888370407954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,3584,0.03494755427042643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,65536,768,0.025289777252409194
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,65536,512,0.014715555641386243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,768,0.14688177903493246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,768,0.01367555558681488
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,65536,256,0.00923644420173433
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,512,0.012727110750145383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,65536,512,0.021743110484547083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,512,0.14033867253197566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,65536,128,0.007781333393520779
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,65536,256,0.01926933394538032
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,256,0.1450808842976888
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,256,0.011424889167149862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,65536,128,0.01787911189926995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,65536,64,0.007312888900438945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,128,0.01126488877667321
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,65536,32,0.0069022224181228215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,65536,128,0.1391413344277276
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,65536,64,0.020909332566791113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,65536,32,0.021534222695562575
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,65536,2048,0.04043289025624593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,16384,16384,0.09063288900587295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,16384,65536,0.26860178841484916
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,16384,65536,0.3441644509633382
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,1536,0.01918577816751268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,16384,12288,0.07043555710050795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,16384,0.04423377911249796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,65536,0.1554044485092163
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,65536,1024,0.015468445089128284
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,16384,12288,0.057724442746904164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,16384,10240,0.057481778992546924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,16384,0.14516177442338732
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,12288,0.03623644510904948
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,16384,10240,0.05036177900102404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,16384,8192,0.05025688807169596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,10240,0.029951999584833782
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,12288,0.1422871086332533
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,10240,0.14620888233184814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,16384,7168,0.04268977708286709
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,8192,0.026345777842733595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,16384,6144,0.03664711117744446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,8192,0.14104621940188938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,16384,7168,0.03918755716747708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,7168,0.022925333844290838
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,7168,0.14504178365071616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,16384,6144,0.03490755624241299
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,6144,0.021077333225144282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,6144,0.14172977871365017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,16384,5120,0.031523555517196655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,16384,16384,0.0766400032573276
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,16384,4096,0.026854222019513447
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,5120,0.14758933915032282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,5120,0.018959111637539335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,65536,0.28940799501207143
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,16384,4096,0.0277804434299469
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,16384,3584,0.02351911034848955
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,4096,0.1464951170815362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,16384,3072,0.020946666598320007
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,3584,0.01531377765867445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,16384,3584,0.02517866591612498
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,16384,3072,0.023511999183230933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,3072,0.013608888619475894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,16384,2560,0.017514665921529133
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,16384,8192,0.043933331966400146
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,16384,2560,0.02128533356719547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,2560,0.14571466710832384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,16384,2048,0.014845333165592618
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,16384,5120,0.03166666626930237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,16384,2048,0.019620445039537217
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,16384,1536,0.012053333222866058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,2048,0.011032000184059143
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,16384,1536,0.017872888180944655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,1536,0.14087200164794922
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,16384,1024,0.009327111144860586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,4096,0.015796444482273526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,1536,0.009737778041097853
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,16384,1024,0.01683644453684489
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,16384,768,0.00815822184085846
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,3072,0.14258933067321777
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,768,0.14223466979132757
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,768,0.0075804442167282104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,16384,512,0.006588444527652528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,2048,0.14883822864956325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,16384,512,0.014075555735164218
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,512,0.14636177486843532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,16384,256,0.004554666578769684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,512,0.006996444529957241
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,16384,256,0.012741333080662621
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,256,0.14093600379096136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,16384,128,0.0038248888320393036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,256,0.006602666858169768
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,16384,768,0.016015110744370353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,16384,128,0.013060444758998023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,128,0.13880977365705702
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,128,0.00699733363257514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,16384,32,0.003504000190231535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,2560,0.012240889171759287
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,16384,64,0.013396444420019785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,16384,32,0.014367111855083041
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,12288,65536,0.27071910434299046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,12288,65536,0.21590844790140787
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,12288,16384,0.07114666700363159
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,12288,16384,0.06376889016893175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,1024,0.1441884438196818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,16384,0.03977600071165297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,12288,12288,0.05612888601091173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,16384,64,0.003504000190231535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,12288,0.14471022288004556
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,12288,0.03064799971050686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,12288,10240,0.04618844389915466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,65536,0.2583084371354845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,12288,10240,0.04236622320281135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,12288,8192,0.040001776483323835
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,10240,0.14473599857754177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,10240,0.02741688821050856
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,65536,0.13591021961636013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,12288,8192,0.037477334340413414
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,12288,7168,0.0339937773015764
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,8192,0.022640888889630634
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,8192,0.14376978079477945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,16384,1024,0.00832799987660514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,12288,7168,0.033358222908443876
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,16384,0.1488142278459337
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,7168,0.020785777105225455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,7168,0.14269422160254583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,12288,5120,0.025651555922296312
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,12288,12288,0.04818844464090136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,6144,0.1423866616355048
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,6144,0.018608889646000333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,12288,4096,0.02194044490655263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,5120,0.14122754997677273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,5120,0.016695999436908297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,12288,4096,0.023817777633666992
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,12288,3584,0.01923111081123352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,4096,0.14236977365281847
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,12288,3584,0.021968000464969214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,3584,0.13876444763607448
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,12288,3072,0.0170462214284473
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,3584,0.01350933313369751
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,12288,3072,0.020768889122539096
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,3072,0.14681955178578696
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,12288,2560,0.014383110735151501
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,16384,3584,0.14074933528900146
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,12288,2560,0.019504000743230183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,2560,0.1405120028389825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,12288,2048,0.012639111114872826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,2560,0.011643555429246692
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,12288,6144,0.030227555168999568
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,12288,2048,0.018192888961897958
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,12288,1536,0.010375111467308467
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,2048,0.14702488316429987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,2048,0.010127999716334874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,12288,5120,0.02684355609946781
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,12288,1024,0.008536888493431939
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,12288,1536,0.016813332835833233
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,1536,0.147361781862047
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,1536,0.009011555876996782
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,12288,768,0.007369777394665613
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,12288,1024,0.015397333436542086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,4096,0.014876445134480795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,1024,0.008126222425036961
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,12288,512,0.005548444473081165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,12288,768,0.014127110441525778
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,768,0.14185244507259792
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,768,0.007298666569921706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,12288,256,0.004214222232500712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,12288,512,0.013431111143694984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,512,0.0069751110341813825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,512,0.14477154943678114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,3072,0.012831110921170978
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,12288,128,0.0038728887836138406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,12288,256,0.01274222218328052
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,256,0.14571289221445718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,12288,64,0.003232888877391815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,12288,128,0.012712889247470431
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,128,0.006647999915811751
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,128,0.1384577751159668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,12288,32,0.0035377778112888336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,12288,6144,0.0298942228158315
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,12288,64,0.012084444363911947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,12288,32,0.0136817776494556
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,10240,65536,0.204492449760437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,10240,16384,0.054047111007902354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,10240,65536,0.20827466911739775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,10240,16384,0.06131644381417168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,65536,0.13051199913024902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,16384,0.03773777683575948
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,10240,12288,0.042671110894944936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,12288,1024,0.140883551703559
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,16384,0.14211911625332302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,10240,10240,0.03567466802067227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,10240,12288,0.04656711220741272
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,12288,0.029814223448435467
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,10240,8192,0.030757334497239854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,12288,0.14279022481706408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,10240,0.02606133288807339
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,10240,0.1398906707763672
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,12288,256,0.006586666736337874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,10240,7168,0.02702666653527154
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,8192,0.14226667086283365
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,65536,0.2484960026211209
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,10240,7168,0.032519999477598406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,10240,6144,0.023577777875794306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,7168,0.1397955550087823
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,7168,0.019430221782790292
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,10240,5120,0.01998488936159346
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,6144,0.018039110634062026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,6144,0.14574310514662
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,10240,5120,0.025403555896547105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,10240,4096,0.017865777015686035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,5120,0.016363556186358135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,5120,0.14006400108337402
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,10240,4096,0.022849778334299724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,4096,0.014175110393100314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,4096,0.14402488867441812
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,10240,10240,0.040898667441474065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,10240,3584,0.021546666820844013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,3584,0.14383378293779162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,10240,3072,0.013374221821626028
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,3584,0.013723555538389417
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,10240,8192,0.03629600008328756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,8192,0.022184888521830242
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,10240,2560,0.012048000262843238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,10240,3072,0.020344000723626878
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,3072,0.1407511101828681
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,3072,0.012306667036480375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,10240,2048,0.010363555616802638
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,2560,0.011349333657158745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,2560,0.14593333668178982
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,10240,1536,0.008667555948098501
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,10240,2048,0.018107554978794523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,2048,0.1391351090537177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,2048,0.009360888765917884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,10240,6144,0.029559112257427637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,10240,1536,0.016220443778567843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,1536,0.008389333056079017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,1536,0.14538844426472983
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,10240,768,0.005897777775923411
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,10240,1024,0.014816888504558139
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,1024,0.007896000312434303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,10240,512,0.004543111142185
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,10240,768,0.013794666363133324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,768,0.007254222200976477
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,768,0.14043822553422716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,10240,256,0.0039013334446483185
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,10240,512,0.013415111435784234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,512,0.00686488880051507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,10240,3584,0.015152000718646579
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,512,0.14973689450158015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,10240,128,0.0035004446076022256
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,10240,256,0.012723555167516073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,256,0.006590222318967183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,10240,128,0.01239466667175293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,128,0.13429333104027644
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,10240,128,0.006583111153708563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,10240,64,0.0031991110493739447
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,10240,32,0.0035111109415690103
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,10240,64,0.01275199982855055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,10240,32,0.013396444420019785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,10240,2560,0.019182221757041085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,8192,65536,0.18242755201127794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,10240,1024,0.007352888584136963
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,8192,65536,0.16555466916826037
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,1024,0.1431822246975369
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,65536,0.07839555872811212
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,8192,16384,0.05067111055056254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,8192,12288,0.0377075539694892
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,16384,0.14163822597927517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,16384,0.025405334101782903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,8192,12288,0.038534220721986555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,8192,10240,0.03119200136926439
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,12288,0.01962844365172916
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,12288,0.1488986677593655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,10240,256,0.14602577686309814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,8192,8192,0.02639377779430813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,65536,0.2308871216244168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,10240,0.017125333348910015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,10240,0.14145067003038195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,8192,8192,0.030722667773564655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,8192,0.14595645003848606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,8192,0.014942222171359591
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,8192,6144,0.020438222421540152
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,7168,0.1449822187423706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,7168,0.013884444203641681
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,8192,16384,0.047967112726635404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,8192,6144,0.024269334144062463
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,8192,5120,0.017380444539917838
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,6144,0.1470551093419393
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,8192,5120,0.022054221895005967
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,5120,0.011938666303952536
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,5120,0.14322133858998617
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,8192,4096,0.02050933407412635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,4096,0.010793777803579966
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,4096,0.14350844754113093
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,8192,3584,0.013359111216333179
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,8192,10240,0.034232887956831194
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,8192,3584,0.019323555959595572
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,3584,0.009471110999584198
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,3584,0.1411235597398546
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,8192,3072,0.012073777616024017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,8192,2560,0.010659555594126383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,8192,3072,0.01847111185391744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,3072,0.14827111032274035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,3072,0.009027555584907532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,8192,7168,0.027448000179396734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,8192,2048,0.009249777429633671
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,2560,0.14039733674791124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,8192,2048,0.01646399994691213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,2048,0.007708444363541073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,8192,1536,0.007845333053006066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,6144,0.012811555630630918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,8192,1536,0.014697778556081982
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,1536,0.1422791083653768
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,8192,1024,0.006574222197135289
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,1536,0.007270221908887227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,8192,1024,0.014087999860445658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,8192,768,0.005215999980767568
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,1024,0.006614222294754452
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,1024,0.14004266262054443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,768,0.006385777973466449
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,8192,768,0.013759111364682516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,768,0.14818578296237522
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,8192,512,0.012792000340090858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,8192,7168,0.023217777411142986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,512,0.14520711368984646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,8192,256,0.003547555456558863
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,512,0.006256000035338932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,8192,256,0.012087999946541257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,2560,0.008368888662921058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,256,0.14016000429789224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,256,0.005939555664857228
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,2048,0.14282133844163683
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,8192,128,0.012102222277058495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,8192,64,0.0031600000543726813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,8192,128,0.13283556037478977
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,8192,32,0.0031431110368834603
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,8192,64,0.012466666599114736
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,8192,4096,0.015187554889255099
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,8192,32,0.012703999876976013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,7168,65536,0.1615635553995768
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,7168,16384,0.043472889396879405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,65536,0.07609244187672932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,7168,16384,0.04916888806555006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,8192,2560,0.017265778448846605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,16384,0.1399431096182929
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,16384,0.02366488840844896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,7168,12288,0.03492888808250427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,8192,128,0.0031902222997612427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,65536,0.22857332229614258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,7168,12288,0.037567999627855085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,7168,10240,0.029097778929604426
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,8192,128,0.005952000204059813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,12288,0.14539466963873968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,7168,8192,0.025035555164019268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,7168,10240,0.03311466508441501
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,10240,0.01648355523745219
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,10240,0.14587467246585423
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,7168,65536,0.16251288519965276
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,7168,7168,0.02161955502298143
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,8192,0.014495111174053617
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,7168,8192,0.02959733208020528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,7168,6144,0.01905155513021681
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,7168,7168,0.02591200007332696
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,7168,0.013437333206335703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,7168,0.1445644431644016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,8192,512,0.0042160000238153665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,7168,6144,0.0243511117166943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,6144,0.012437333663304647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,6144,0.14152533478207058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,7168,4096,0.014539556370841132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,7168,5120,0.02197866638501485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,5120,0.1468986670176188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,5120,0.011411555939250521
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,7168,3584,0.012492444780137805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,7168,4096,0.02022133270899455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,4096,0.010048000348938836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,4096,0.1406035555733575
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,7168,3072,0.010992888775136737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,7168,3584,0.01920622256067064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,3584,0.00906755526860555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,3584,0.14331732855902776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,12288,0.018799111247062683
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,7168,2560,0.009860444400045607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,7168,3072,0.0185244447655148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,3072,0.14157599873012966
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,7168,2048,0.008603555460770925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,7168,2560,0.01741333305835724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,2560,0.007975111405054728
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,7168,2048,0.016484444340070088
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,8192,0.1401911046769884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,7168,1536,0.007532444265153673
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,2048,0.0074551113777690465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,7168,1536,0.014535110857751636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,1536,0.14354311095343694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,7168,1024,0.005888888819350137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,7168,5120,0.016343111793200176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,7168,1024,0.013990221752060784
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,1024,0.00646666685740153
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,1024,0.1401724418004354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,7168,768,0.013082666529549493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,768,0.1397706667582194
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,7168,512,0.0041599998043643106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,768,0.006232000059551663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,7168,512,0.01272266689274046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,3072,0.008658666577604082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,7168,256,0.0035084444615576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,512,0.006226666685607698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,7168,256,0.012196444802814059
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,2048,0.14788088533613417
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,256,0.14085777600606283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,7168,128,0.003535110917356279
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,256,0.005955555372767978
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,1536,0.0069822221994400024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,7168,128,0.012088889049159156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,7168,64,0.0028595555987623003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,128,0.1370133293999566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,7168,128,0.005895111295912001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,7168,32,0.0031911111954185697
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,7168,64,0.012410666379663678
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,7168,768,0.004982222285535601
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,7168,32,0.011742221812407175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,6144,65536,0.1328479978773329
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,6144,16384,0.04159733321931627
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,6144,65536,0.1561608844333225
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,6144,16384,0.04795822170045641
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,6144,12288,0.03386577632692125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,65536,0.07388799720340304
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,16384,0.022613333331214056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,16384,0.14495821793874106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,6144,10240,0.027905778752432928
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,6144,12288,0.03643466697798835
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,12288,0.018191999859280057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,512,0.1452088885837131
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,12288,0.13870222038692898
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,6144,8192,0.02219733264711168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,10240,0.01611288885275523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,7168,2560,0.14224889543321398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,6144,7168,0.019864888654814828
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,65536,0.2262773248884413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,6144,8192,0.02757155564096239
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,8192,0.13896444108751085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,6144,6144,0.017032000753614638
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,6144,7168,0.025828444295459326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,7168,0.14790844917297363
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,6144,5120,0.01535555554760827
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,6144,6144,0.023434667123688593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,6144,0.15234843889872232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,6144,0.012387555506494312
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,6144,5120,0.02127733329931895
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,6144,4096,0.013426666458447775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,5120,0.011043555206722684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,6144,4096,0.01975644462638431
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,4096,0.14237511157989502
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,6144,3584,0.011954666839705573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,4096,0.009377777576446533
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,6144,3584,0.01898755629857381
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,3584,0.1464808914396498
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,6144,3072,0.01105955574247572
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,3584,0.008907555705971187
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,6144,10240,0.03219644559754266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,10240,0.14229599634806314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,6144,2560,0.00960533320903778
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,6144,3072,0.018106665876176622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,8192,0.014122666584120857
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,3072,0.1426524453692966
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,6144,2048,0.008655110994974772
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,7168,0.013148444394270578
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,6144,2560,0.016540444559521146
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,2560,0.00792622235086229
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,2560,0.14527555306752524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,6144,1536,0.007272889216740926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,6144,2048,0.015843555331230164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,2048,0.007273777491516537
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,2048,0.14038044876522487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,6144,1024,0.00554577757914861
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,5120,0.13874755965338811
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,1536,0.0069217777086628815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,6144,768,0.004804444395833545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,6144,1024,0.013730666703648038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,1024,0.14090666506025526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,6144,768,0.013419555293189155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,768,0.14138666788736978
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,768,0.006224888894293044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,6144,512,0.012452444268597497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,512,0.1419688860575358
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,6144,256,0.0035546666218174826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,512,0.005931555396980709
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,6144,256,0.012097777591811286
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,3072,0.00830577810605367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,256,0.13967111375596789
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,6144,128,0.003192000091075897
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,256,0.005910222315125995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,6144,128,0.011736888852384357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,128,0.005913777897755305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,128,0.13649066289265951
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,6144,64,0.0031635556370019913
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,6144,32,0.0031306667046414483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,6144,1536,0.014132445057233175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,6144,1536,0.1400631136364407
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,6144,64,0.012660444610648684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,6144,32,0.012421333127551608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,5120,65536,0.11514488855997722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,6144,1024,0.006594666590293248
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,5120,16384,0.03597866826587253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,5120,65536,0.15149688720703125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,6144,512,0.00388355553150177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,5120,12288,0.029952890343136255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,5120,16384,0.046648889780044556
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,16384,0.14133511649237737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,5120,12288,0.03577955563863119
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,12288,0.01752088963985443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,12288,0.1484417782889472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,5120,10240,0.031726221243540444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,65536,0.22414755821228027
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,10240,0.01574844453069899
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,5120,8192,0.019158222609096102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,5120,8192,0.02788533435927497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,8192,0.14759111404418945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,5120,7168,0.016349333855840895
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,5120,7168,0.02474133339193132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,7168,0.14044888814290366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,5120,6144,0.014448889427714877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,7168,0.01275199982855055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,6144,0.011776888536082374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,6144,0.14484532674153647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,65536,0.07121600045098199
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,5120,5120,0.013068444199032254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,16384,0.021347555849287245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,5120,4096,0.01143200033240848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,5120,0.010677333507272931
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,5120,10240,0.02681955529583825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,5120,0.1440595520867242
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,5120,4096,0.019549333386951022
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,5120,3584,0.010994666980372535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,4096,0.14806666639116076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,4096,0.00868444475862715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,5120,3584,0.01860000027550591
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,5120,3072,0.010311110979980892
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,3584,0.008712000317043727
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,5120,3072,0.017495110630989075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,8192,0.0136817776494556
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,3072,0.1385520034366184
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,5120,2560,0.009224889179070791
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,3072,0.008329778081840938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,5120,2048,0.008219555848174626
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,5120,2560,0.016180444094869826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,2560,0.007613333563009898
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,2560,0.1387057834201389
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,5120,2048,0.014525334040323893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,5120,1536,0.006934222247865465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,2048,0.14453155464596218
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,5120,6144,0.023297778434223596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,5120,1536,0.013771555489963956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,5120,1024,0.005583111196756363
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,1536,0.1425573296017117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,1536,0.006909333169460297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,5120,5120,0.02091288897726271
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,5120,1024,0.01309333327743742
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,5120,768,0.004565333326657613
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,1024,0.143230226304796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,1024,0.006279110908508301
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,5120,512,0.0041777777175108595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,5120,768,0.012774222426944308
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,768,0.006215111249023014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,10240,0.14425244596269396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,768,0.14375821749369302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,3584,0.1427111095852322
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,5120,256,0.003567111161020067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,5120,512,0.012426666915416718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,512,0.005903999838564131
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,5120,128,0.0031999999450312722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,512,0.14891289340125188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,256,0.00591822216908137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,256,0.1415439976586236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,5120,64,0.0028373334142896864
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,5120,128,0.012063999970753988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,5120,32,0.0029884444342719186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,5120,128,0.13309066825442845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,5120,64,0.01237511138121287
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,4096,65536,0.09229599767261082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,5120,32,0.01239111108912362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,4096,16384,0.031192888816197712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,2048,0.007286222444640265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,65536,0.06894310977723864
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,4096,16384,0.03486311104562547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,16384,0.14030310842725965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,16384,0.020591111646758187
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,65536,0.22252710660298666
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,4096,12288,0.027327999472618103
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,4096,10240,0.022461333208613928
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,12288,0.14504178365071616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,12288,0.016941333810488384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,4096,10240,0.024319999747806128
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,10240,0.14404355155097112
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,5120,256,0.012049777640236748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,4096,8192,0.016131555040677387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,10240,0.015277332729763456
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,4096,7168,0.014404444230927361
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,5120,128,0.0059262220230367445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,8192,0.013071111506885953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,8192,0.14828622341156006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,4096,7168,0.020385776956876118
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,4096,6144,0.012713777522246042
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,7168,0.1476693285836114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,7168,0.012256888879670037
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,4096,5120,0.011374221907721626
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,4096,65536,0.110770669248369
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,4096,6144,0.019298666053348117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,6144,0.011063111325105032
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,4096,4096,0.010344888601038191
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,4096,5120,0.017981333865059745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,5120,0.010118222071064843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,5120,0.14077777332729763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,4096,12288,0.025632888078689575
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,4096,3584,0.00905955582857132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,4096,4096,0.016194666425387066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,4096,0.008619555168681675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,4096,0.13948710759480795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,4096,3072,0.008617777791288164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,4096,3584,0.016373333003785875
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,4096,3072,0.01482399966981676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,3072,0.1430328951941596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,4096,2560,0.007800888684060838
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,3072,0.007928888830873702
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,4096,2048,0.00693866651919153
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,2560,0.00757422215408749
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,4096,2560,0.014104000396198697
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,2560,0.14246666431427002
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,4096,1536,0.005209777918126848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,4096,2048,0.013741333451535968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,2048,0.007272889216740926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,2048,0.140883551703559
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,4096,8192,0.02197511163022783
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,4096,1024,0.004511111312442356
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,4096,1536,0.01309866706530253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,1536,0.14102578163146973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,1536,0.006923555499977536
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,4096,768,0.0038817777401871155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,4096,1024,0.012758221891191272
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,1024,0.00658044425977601
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,1024,0.14012266529930964
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,4096,512,0.0036764442920684814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,4096,768,0.012719111310111152
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,768,0.14087377654181585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,768,0.006248000181383557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,4096,256,0.0031848889258172777
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,4096,512,0.012376888758606382
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,512,0.14234311050838894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,512,0.005956444475385878
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,4096,128,0.003149333306484752
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,4096,256,0.01203555530971951
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,256,0.1401217778523763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,256,0.00589155571328269
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,4096,64,0.003165333428316646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,3584,0.00832088871134652
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,4096,128,0.011700444751315646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,128,0.13708178202311197
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,4096,64,0.012038222617573209
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3584,65536,0.08247822523117065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,4096,32,0.01205244412024816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3584,65536,0.106404443581899
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,65536,0.06797689199447632
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3584,16384,0.03353244397375319
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,65536,0.2215635511610243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,6144,0.14607732825809056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,16384,0.1418951087527805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,16384,0.020426667398876615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3584,12288,0.0269768883784612
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,12288,0.14444888962639704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,4096,3584,0.14408444033728704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,4096,128,0.0058711109062035876
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,4096,32,0.003157333367400699
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3584,10240,0.02403555479314592
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,10240,0.015119999647140503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,10240,0.14004977544148764
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3584,8192,0.01866133345497979
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3584,8192,0.02162577708562215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,8192,0.14082133769989014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,8192,0.012665777570671506
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3584,16384,0.02865155537923177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3584,7168,0.020012444920010038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,7168,0.011915555430783166
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,7168,0.14419821898142496
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3584,6144,0.01276177747382058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3584,6144,0.018935999936527677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,6144,0.010815111299355825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3584,5120,0.011013333168294696
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,6144,0.14324266380733913
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3584,12288,0.02329155637158288
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3584,4096,0.010045333041085137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3584,5120,0.01759466694460975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,5120,0.009999111294746399
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,5120,0.1435448858473036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3584,3584,0.008719111482302347
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,4096,0.008566222257084316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,12288,0.016686222619480558
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3584,3584,0.015743111570676167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3584,3072,0.008264888491895463
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,3584,0.14155644840664333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3584,3072,0.01482399966981676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,3072,0.1423226727379693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3584,7168,0.013759111364682516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,3072,0.008009778128729926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3584,2560,0.007253333098358578
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3584,2560,0.01384800010257297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3584,2048,0.005979555762476391
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3584,2048,0.013382222089502545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,2048,0.14788799815707737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3584,1536,0.004905777672926585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,2048,0.007273777491516537
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,4096,0.13996533552805582
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3584,1536,0.012811555630630918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3584,1024,0.0041911109454101985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,1536,0.006633777585294511
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,1536,0.13914044698079428
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,3584,0.00828888929552502
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3584,768,0.0038364442686239877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3584,1024,0.0124177775449223
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,1024,0.006344888773229387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,1024,0.14075556066301134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3584,512,0.0037662221325768363
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3584,768,0.012432888978057437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,768,0.006248888870080312
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3584,512,0.012279111478063794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,2560,0.007600000335110559
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3584,256,0.0034844444857703317
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,512,0.005896889087226655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,2560,0.14272977246178523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3584,128,0.002887111157178879
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,256,0.005882666756709416
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3584,256,0.011767999993430244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,256,0.13830933305952284
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3584,128,0.01143200033240848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,3584,128,0.0059004442559348206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,128,0.1327528953552246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3584,64,0.01204622205760744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3072,65536,0.0725475549697876
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3584,10240,0.021036444438828364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3072,65536,0.1063199970457289
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,768,0.14935199419657388
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,65536,0.221015108956231
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3072,16384,0.023003555006451074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,3584,512,0.14056621657477483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3072,16384,0.03254400028122796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3072,12288,0.019310222731696237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,16384,0.14807644155290392
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3584,64,0.002954666606254048
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3072,12288,0.02572088936964671
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3584,32,0.0031511110977994073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,12288,0.1476480033662584
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3072,10240,0.016562667157914903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,12288,0.016547555724779766
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3584,32,0.012067555553383298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3072,10240,0.023287110858493384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3584,4096,0.01608888970481025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3072,8192,0.01445777714252472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,10240,0.14307288328806558
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,10240,0.014817777607176038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3072,7168,0.013163555827405719
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3072,8192,0.02122666603989071
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,8192,0.14261421892378065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3072,7168,0.01996977792845832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,7168,0.14199199941423205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3072,6144,0.018536888890796237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,65536,0.06697599755393134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,6144,0.14232177204555935
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3072,5120,0.01037599974208408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,6144,0.010354667074150508
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3072,5120,0.017314665847354464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,5120,0.14353956116570368
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3072,4096,0.009698666632175446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,16384,0.019913777709007263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3072,4096,0.016030222177505493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,4096,0.008608889248636033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,4096,0.14352088504367405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3072,3584,0.008608889248636033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3072,3584,0.01475288967291514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,3584,0.008291555775536431
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,3584,0.14233599768744573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3072,3072,0.007970666719807519
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3072,3072,0.014499555031458536
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,3072,0.00797155582242542
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3072,2560,0.007291555404663086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,3072,0.14174666669633654
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3072,6144,0.01203555530971951
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3072,2560,0.013799111048380533
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,2560,0.0075937774446275495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,2560,0.14264267020755345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3072,2048,0.013444444371594323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,2048,0.14286222722795275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3072,1536,0.00488800017370118
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3072,1536,0.01291911138428582
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,1536,0.1399635606341892
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,5120,0.00962755580743154
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3072,1024,0.00416355538699362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,1536,0.006624889042642381
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3072,1024,0.012440889245933957
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,1024,0.139847993850708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3072,768,0.0038622220357259116
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,1024,0.006412444429265128
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3072,768,0.012432888978057437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,768,0.006281777802440856
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,768,0.14600443840026855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3072,512,0.0035982223020659555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3072,512,0.012263110942310758
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,512,0.005894222193294101
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,8192,0.012412444584899478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3072,256,0.0032239999208185407
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,512,0.14580444494883218
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3072,128,0.0031759999692440033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3072,2048,0.0062871111763848206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3072,256,0.011692444483439127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,256,0.14487377802530924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3072,128,0.011821333732869891
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3072,64,0.0028693332440323303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,3072,32,0.002805333377586471
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,128,0.005879111174080107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,2048,0.007258666886223688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2560,65536,0.06337066491444905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3072,32,0.01179377817445331
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2560,65536,0.10203199916415745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2560,16384,0.02197511163022783
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,65536,0.0655768911043803
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2560,16384,0.031249778138266668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,16384,0.01976444489426083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,65536,0.220268448193868
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2560,12288,0.02530044482813941
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,12288,0.13911555873023138
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2560,10240,0.016202666693263583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,12288,0.0162444445821974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2560,10240,0.022898667388492163
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,7168,0.011383111278216044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2560,8192,0.014498665928840637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,10240,0.141595549053616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,10240,0.013953778478834363
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,3072,128,0.1407368845409817
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2560,8192,0.020632000433074105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2560,7168,0.012842666771676807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,3072,64,0.011789333489206104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,8192,0.14553866121504042
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,8192,0.011680000358157687
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2560,6144,0.011558222273985544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2560,7168,0.019294222195943195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,7168,0.011367110742463006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,7168,0.14453332953982884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2560,5120,0.010273777776294285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,6144,0.1447626617219713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,6144,0.010371555884679159
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2560,5120,0.016523554921150208
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,5120,0.1459840006298489
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2560,12288,0.01808533403608534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,16384,0.14239911238352457
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2560,4096,0.015783111254374187
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,4096,0.008656000097592672
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2560,3584,0.00811111099190182
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2560,3584,0.014458666245142618
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,3584,0.14531732930077448
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2560,3072,0.007637333538797166
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,3072,256,0.005673777725961473
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,3584,0.008282666405042013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2560,3072,0.014177777700954013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,3072,0.007954667011896769
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,3072,0.14284533924526638
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2560,2560,0.013741333451535968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,2560,0.1429493294821845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2560,2048,0.0058622223635514574
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,2560,0.007617777420414819
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2560,2048,0.013447999954223633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2560,4096,0.009362666971153682
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,5120,0.009687110781669617
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2560,1536,0.004931555440028508
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,2048,0.0069644442862934535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2560,1536,0.012746666868527731
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,4096,0.13935200373331705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,1536,0.14345599545372859
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,1536,0.006672888994216919
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2560,1024,0.012706666356987424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,1024,0.006265777680608962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,1024,0.1447626617219713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2560,768,0.012430222498046027
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,768,0.14553511142730713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2560,512,0.0035048888789282907
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2560,2560,0.006919999917348226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,768,0.006222222414281633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2560,512,0.01203733351495531
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2560,256,0.003165333428316646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2560,6144,0.01811733345190684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2560,256,0.011696889168686338
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,256,0.13956088489956325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2560,128,0.0029271110478374693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,256,0.005715555614895291
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2560,128,0.011365333365069495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,128,0.13392889499664307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2560,1024,0.0042204442951414315
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2560,64,0.002879111096262932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2560,32,0.0028106667515304354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,128,0.005877333382765452
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2560,64,0.011363555159833698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2560,768,0.004027555386225383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2048,65536,0.05200799968507555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2560,32,0.011543110840850405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2048,16384,0.01854222185081906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2048,65536,0.10005688667297363
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,65536,0.06410844458474053
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2048,16384,0.030761778354644775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2048,12288,0.015431111057599386
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,16384,0.01924977699915568
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,65536,0.2228684425354004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,16384,0.14812089337242976
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,512,0.1448800033993191
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2048,10240,0.014640889234013028
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,12288,0.015620445211728414
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2048,8192,0.013094222380055321
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2048,10240,0.022325333621766832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,10240,0.013392888837390475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,10240,0.14594311184353298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2048,7168,0.011758222348160215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,2560,2048,0.14143555694156223
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,8192,0.01161688897344801
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,8192,0.1479733387629191
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2048,6144,0.011222222612963783
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2048,7168,0.01926577753490872
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,7168,0.14093955357869467
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2048,6144,0.01717155509524875
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,6144,0.14583110809326172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,6144,0.010027555955780877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2048,5120,0.01649688846535153
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,5120,0.14154577255249023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2048,4096,0.010314666562610203
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,5120,0.00961866643693712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2048,4096,0.014841778410805596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,2560,512,0.005888888819350137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,4096,0.14604177739885119
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,12288,0.13967466354370117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2048,3584,0.009335999687512716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,4096,0.008613333106040955
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2048,3072,0.008637333081828224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2048,3584,0.014816888504558139
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,3584,0.008291555775536431
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,3584,0.13912533389197454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2048,2560,0.007231111327807109
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2048,8192,0.02051555613676707
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2048,3072,0.014140443669425117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,3072,0.007655111451943715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,3072,0.14502667056189641
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2048,2048,0.006566222343179915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2048,2560,0.013735111388895245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,2560,0.007559110720952352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,2560,0.13993955983055964
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,7168,0.010693333215183683
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2048,1536,0.004744889007674323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2048,2048,0.013442666994200813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,2048,0.1424693399005466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,2048,0.0070773329999711775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2048,1024,0.0041955556306574075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2048,1536,0.012748444245921241
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,1536,0.14028178320990667
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2048,768,0.003847111016511917
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2048,1024,0.012429333395428128
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,1024,0.00638133328821924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,1024,0.14504088295830622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2048,512,0.003549333247873518
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2048,768,0.012392000191741519
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,768,0.1405715545018514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2048,512,0.012106666962305704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2048,256,0.003152888889114062
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,512,0.005897777775923411
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2048,12288,0.024682665864626568
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2048,256,0.011751111182901593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,256,0.005620444400442972
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2048,128,0.0028533333291610083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2048,128,0.011416888899273343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,128,0.13465778032938638
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2048,64,0.0028071111689011254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,128,0.005569777968857024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2048,32,0.0031724443866146933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2048,64,0.011742221812407175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,2048,5120,0.012445333103338877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,2048,32,0.011700444751315646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,1536,0.006969778074158563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1536,65536,0.09686222341325547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1536,16384,0.017955554856194388
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,2048,768,0.006231110956933763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,65536,0.21997955110338
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1536,16384,0.02994489007525974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,512,0.14135555426279703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1536,12288,0.01479288935661316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,16384,0.018744000130229525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,2048,256,0.1413546668158637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,12288,0.014105778601434497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,12288,0.14895110660129124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1536,10240,0.013432888521088494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1536,10240,0.022292445103327434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,10240,0.013083555632167392
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,10240,0.14818399482303196
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1536,8192,0.01921066641807556
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,8192,0.1475600004196167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1536,65536,0.039023998710844256
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,8192,0.011298666397730509
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1536,7168,0.018255111243989732
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,65536,0.0628346668349372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,7168,0.14587467246585423
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1536,6144,0.010682666467295753
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,7168,0.010675555302037133
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1536,6144,0.016768000192112394
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1536,5120,0.010046222143703038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,6144,0.010024888647927178
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,16384,0.14130311542087132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1536,5120,0.016173332929611206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1536,12288,0.024485333098305598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,5120,0.14479466279347739
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1536,4096,0.009298666483826106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,4096,0.008332444561852349
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1536,4096,0.01499466598033905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,4096,0.14056621657477483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1536,8192,0.012423999607563019
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1536,3072,0.006913777854707506
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1536,3584,0.014440889159838358
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,3584,0.007982222570313348
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,3584,0.14548356003231472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1536,2560,0.005933333188295364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1536,7168,0.011707555916574268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,3072,0.00776800016562144
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,3072,0.14708622296651205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1536,2048,0.0052435555391841466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1536,2560,0.013739555246300168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,2560,0.007264888948864407
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,2560,0.14890044265323216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1536,1536,0.004567999806669024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1536,2048,0.013140444126394061
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,2048,0.006959999601046245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,2048,0.14711556169721815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1536,1024,0.004168888760937585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1536,1536,0.012753778033786349
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,1536,0.006637333167923822
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,1536,0.14632711145612928
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1536,768,0.003620444486538569
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1536,1024,0.01240888900227017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,1024,0.0063288890653186375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1536,768,0.012087110843923358
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,5120,0.009374221993817223
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1536,512,0.0032142222755485107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,768,0.006173333360089197
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1536,512,0.01203733351495531
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1536,3584,0.007250666618347168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,512,0.005889777921968036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1536,256,0.00318755561278926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,512,0.1443822251425849
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1536,256,0.011697777443461947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,256,0.005902222047249476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,256,0.14232533507876927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1536,128,0.002830222249031067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1536,3072,0.014104889498816596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1536,64,0.002826666666401757
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1536,128,0.011411555939250521
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,1536,128,0.005604444278611078
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,128,0.13298488987816706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1024,65536,0.02940444482697381
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1536,64,0.011696889168686338
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1536,32,0.011708444191349877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1024,65536,0.09344977802700466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,65536,0.05951466825273302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,6144,0.14477245012919107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1024,16384,0.013943110903104147
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,65536,0.219650665918986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,1024,0.14764977826012507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1024,16384,0.029895110262764826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,16384,0.017826666434605915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,1536,768,0.14448266559176975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1024,12288,0.023453333311610754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1024,10240,0.011575111084514193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,12288,0.1398551066716512
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1024,10240,0.02128444446457757
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,10240,0.012983111043771109
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,10240,0.13985244433085123
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1024,8192,0.013363555901580386
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1024,8192,0.01887822151184082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,8192,0.011399110986126794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,8192,0.13935644096798366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1024,7168,0.012072000238630505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1536,32,0.0028213332924577924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1024,6144,0.010699555277824402
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,7168,0.010632889138327705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,7168,0.14459733168284097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1024,6144,0.016529776983790927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,6144,0.010001777774757808
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1024,5120,0.00965244405799442
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1024,5120,0.015779554843902588
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,5120,0.1501164436340332
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1024,12288,0.012379555238617791
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1024,4096,0.0086986662613021
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,16384,0.14383555783165827
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,5120,0.009287999735938178
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1024,3584,0.007237333390447829
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1024,4096,0.0147706667582194
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,12288,0.014102222190962898
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,4096,0.14728532897101507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1024,3072,0.006573333508438534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1024,3584,0.014483556151390076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,3584,0.008250666989220513
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,3584,0.1418044434653388
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1024,2560,0.005946666830115848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,3072,0.007617777420414819
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1024,2560,0.013422222601042854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,2560,0.14244622654385036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1024,7168,0.017888888716697693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1024,2048,0.006239999913507038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1024,2048,0.013085333009560903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,2048,0.14097155465020075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1024,1536,0.005581333405441708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,2048,0.007034666836261749
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1024,1536,0.01275111072593265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,1536,0.1480568912294176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1024,1024,0.004561777744028303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,1536,0.006657777975002925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,4096,0.008344888687133789
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1024,768,0.0038604442444112566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,1024,0.00628355559375551
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,1024,0.14042311244540745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1024,3072,0.01404266721672482
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,768,0.006108444597986009
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1024,512,0.003830222205983268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,768,0.1405715545018514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,3072,0.14065333207448324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1024,256,0.003142222141226133
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,512,0.006088888893524806
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,512,0.1463164488474528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1024,256,0.011740444435013665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,2560,0.00758755538198683
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,256,0.005622222191757626
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1024,128,0.011425777441925473
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,6144,0.13927377594841853
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1024,64,0.0028497777465316984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,128,0.13490933842129177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1024,32,0.0028799999919202593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,1024,128,0.005912888795137405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,768,65536,0.030357334348890517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1024,64,0.011688888900809817
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1024,32,0.011420444481902652
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,768,65536,0.09121777613957723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,768,16384,0.011733333269755045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1024,1024,0.012063999970753988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,65536,0.21827022234598795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,768,16384,0.029162665208180744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1024,768,0.01240888900227017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,16384,0.14885777897304958
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,768,12288,0.010598222414652506
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,768,12288,0.023337778117921617
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,1024,512,0.012075555821259817
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,12288,0.14720000161064997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,768,10240,0.009674666656388177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,1024,128,0.0028524444335036804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,12288,0.014116444521480136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,768,8192,0.009740444521109263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,1024,256,0.14203555054134792
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,10240,0.1399386723836263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,10240,0.012824888858530255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,768,7168,0.009305777649084726
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,768,8192,0.018920888503392536
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,8192,0.1414346694946289
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,768,7168,0.017901332841979135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,768,6144,0.007977777885066139
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,7168,0.14649777942233616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,768,6144,0.016806221670574613
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,6144,0.14736444420284694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,65536,0.05818844503826565
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,768,5120,0.008783111141787635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,6144,0.010000000397364298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,768,5120,0.015623999966515435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,16384,0.01685422162214915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,768,4096,0.007657777931955125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,5120,0.1454346709781223
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,4096,0.008378666308191087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,4096,0.14597154988182917
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,768,3584,0.007972444097201029
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,768,10240,0.020296888218985665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,768,3584,0.0144177774588267
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,3584,0.14508888456556532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,768,3072,0.006285333385070165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,3584,0.007975999679830339
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,8192,0.011344888971911537
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,768,3072,0.014119999276267158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,768,2560,0.005671111245950063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,3072,0.14515554904937744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,3072,0.0076248885856734375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,7168,0.010657777388890585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,768,2048,0.00527022240890397
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,768,2560,0.013776000175211163
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,2560,0.007312888900438945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,2560,0.14081154929267034
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,768,1536,0.004524444540341695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,768,2048,0.01347555551264021
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,2048,0.007297777467303806
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,768,1024,0.0038764443662431505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,1536,0.006643555644485686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,768,1024,0.012420444852775998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,5120,0.009311999711725447
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,768,768,0.0038408889538711975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,1024,0.006269333263238271
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,768,4096,0.014780445231331719
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,768,768,0.01236977759334776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,768,512,0.0035111109415690103
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,768,0.00620977787507905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,768,0.14592444896697998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,768,512,0.012099555797047086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,512,0.005920889063013925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,512,0.13892711533440485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,768,256,0.011883555187119378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,256,0.13917156060536703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,768,128,0.002809777855873108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,256,0.00574311117331187
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,2048,0.147816883193122
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,768,128,0.011696000066068439
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,768,64,0.0028008888992998335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,128,0.13611200120713976
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,1536,0.14661155806647405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,768,32,0.002838222309947014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,512,65536,0.022181333767043218
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,768,64,0.011688888900809817
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,768,32,0.011420444481902652
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,512,16384,0.012695999609099494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,512,65536,0.08937155538135105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,65536,0.05780888928307427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,65536,0.22023910946316191
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,512,16384,0.027850667635599773
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,512,12288,0.010316444767846001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,16384,0.14398845036824545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,512,12288,0.022682666778564453
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,12288,0.14490755399068198
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,512,10240,0.010459555519951714
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,768,256,0.003195555466744635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,512,10240,0.020579554968410067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,10240,0.012726222475369772
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,512,8192,0.009329777624871995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,512,8192,0.018869333797030978
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,8192,0.14840088950263128
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,512,7168,0.008964444200197855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,768,1536,0.012703110774358114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,8192,0.011338666909270816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,768,128,0.005635555419656966
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,512,6144,0.007627555893527136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,7168,0.010654222634103564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,7168,0.13891643948025173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,768,1024,0.14301155673133
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,512,6144,0.016329778565300834
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,512,5120,0.008276444342401292
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,6144,0.14128888977898493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,6144,0.010016889207892949
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,512,5120,0.01531644496652815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,16384,0.01687466601530711
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,5120,0.00941955546538035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,512,4096,0.007605333295133378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,12288,0.014131555954615274
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,512,4096,0.014774221513006421
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,4096,0.008454221818182204
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,512,3584,0.006948444578382704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,4096,0.14040266142951116
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,512,3072,0.006535999890830781
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,512,3584,0.01441422270403968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,10240,0.1399431096182929
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,3584,0.1483466625213623
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,512,2560,0.0059057776298787855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,3072,0.007620444728268518
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,3072,0.1392266617880927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,512,2560,0.013419555293189155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,512,2048,0.0052373334765434265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,2560,0.007446222007274628
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,2560,0.14118399884965685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,512,7168,0.017525333497259352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,512,2048,0.01277599980433782
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,2048,0.007290666302045186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,2048,0.14073510964711508
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,512,1536,0.01240799989965227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,1536,0.00685422205262714
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,512,1024,0.0038808890514903595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,1536,0.14263821972741023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,5120,0.14378844367133245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,1024,0.006287999865081575
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,512,768,0.0035457776652442086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,512,768,0.012094222009181976
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,768,0.14109867148929173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,512,512,0.0034924443397257063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,3584,0.008078222473462423
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,512,3072,0.014091556270917257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,512,512,0.012044444680213928
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,512,256,0.0031991110493739447
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,512,0.1438408825132582
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,512,0.005946666830115848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,512,256,0.011762667033407422
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,256,0.0058853332367208265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,512,128,0.00282222218811512
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,256,0.14350133472018772
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,512,1536,0.004541333350870344
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,512,64,0.0028213332924577924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,512,128,0.011759999725553723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,128,0.0059262220230367445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,128,0.1324559979968601
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,512,32,0.002842666581273079
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,256,65536,0.016159999701711867
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,512,64,0.011361777782440186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,512,32,0.01110400011142095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,256,16384,0.009750222166379293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,256,65536,0.08459911081526016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,512,1024,0.012347555822796293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,65536,0.05269155568546719
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,512,1024,0.1473591062757704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,65536,0.21798843807644316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,256,12288,0.00868800034125646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,256,16384,0.028008000718222723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,16384,0.016872000363137987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,256,10240,0.00903733323017756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,12288,0.014099554883109199
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,12288,0.1418737835354275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,256,8192,0.00831733312871721
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,512,768,0.006232000059551663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,256,10240,0.02035999960369534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,10240,0.012798222402731577
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,10240,0.14544088310665554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,256,7168,0.00795999997191959
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,256,8192,0.01868977811601427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,8192,0.14205066363016763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,256,7168,0.017886221408843994
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,256,6144,0.007602666815121968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,7168,0.14679377608829075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,7168,0.010704889065689512
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,256,5120,0.007823111282454597
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,256,6144,0.016535111599498324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,6144,0.01000088867213991
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,6144,0.14198844962649876
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,256,4096,0.007302222152551015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,256,5120,0.015807110402319167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,5120,0.009359111388524374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,5120,0.14658933215671116
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,256,3584,0.0069368887278768755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,256,4096,0.014627556006113688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,4096,0.008296888735559251
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,256,3072,0.006629333313968446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,4096,0.14214221636454263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,256,3584,0.01462399959564209
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,3584,0.008005333443482717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,256,2560,0.006627555522653792
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,256,3072,0.014104000396198697
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,3072,0.140937778684828
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,16384,0.1432373391257392
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,256,12288,0.022302221920755174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,2560,0.007418666448858049
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,2560,0.1421475542916192
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,256,2048,0.012797333300113678
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,2048,0.1422035561667548
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,8192,0.011229332950380115
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,2048,0.00699733363257514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,256,1536,0.012499555945396423
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,1536,0.14794577492607966
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,256,1024,0.003871110992299186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,1536,0.006928000185224745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,256,1024,0.012054222325483957
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,1024,0.13992799652947321
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,256,768,0.003540444291300244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,1024,0.006258666515350342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,3584,0.1463937759399414
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,256,768,0.012085333466529846
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,3072,0.007636444436179266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,256,512,0.003140444349911478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,768,0.1488328907224867
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,256,2048,0.006281777802440856
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,256,256,0.0028044444819291434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,512,0.005922666854328579
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,512,0.1406924459669325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,256,256,0.011740444435013665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,256,0.005767111149099138
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,256,128,0.002528000002106031
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,256,0.14751644929250082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,256,1536,0.004508444418509801
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,256,64,0.0028035555862718155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,256,128,0.011404444773991903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,128,0.005585777676767773
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,256,128,0.13298222753736708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,128,65536,0.015433778365453085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,256,64,0.011749332977665795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,256,32,0.010395555860466426
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,128,65536,0.08504088719685872
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,65536,0.050308446089426674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,128,16384,0.009648000200589498
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,65536,0.2175884379280938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,128,16384,0.026770666241645813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,16384,0.016857778032620747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,128,12288,0.00831822223133511
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,256,768,0.006225777582989798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,128,12288,0.02267644471592373
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,256,2560,0.01334755536582735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,12288,0.14910577403174505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,256,512,0.01205244412024816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,128,8192,0.007283555964628856
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,128,10240,0.020752888586786058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,10240,0.012791111237472959
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,10240,0.14185244507259792
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,128,7168,0.006917333437336816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,8192,0.01108444482088089
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,8192,0.1404542260699802
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,256,32,0.002510222295920054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,128,6144,0.006559999866618051
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,7168,0.010664888554149203
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,7168,0.14909777376386854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,128,6144,0.01646222174167633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,6144,0.010020444790522257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,128,5120,0.0062871111763848206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,6144,0.14046311378479004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,128,5120,0.015134221977657743
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,5120,0.009308444129096137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,5120,0.1401333279079861
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,16384,0.14437333742777506
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,128,4096,0.014899555179807874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,128,3584,0.006299555715587404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,4096,0.0083137775460879
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,128,10240,0.007584888901975419
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,12288,0.014097778333557976
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,128,3584,0.014470222923490735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,128,3072,0.005938666562239329
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,3584,0.14390844768948025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,3584,0.008103110724025303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,128,2560,0.006548444430033366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,128,8192,0.01889955500761668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,128,3072,0.014087999860445658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,3072,0.007643555601437886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,3072,0.14517066213819715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,128,2048,0.006240889016124938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,128,2560,0.013518222504191928
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,128,7168,0.01758666667673323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,2560,0.14639111359914145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,128,1536,0.0041964443193541635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,2048,0.14243200090196398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,2048,0.0069422221018208405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,128,1024,0.003779555360476176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,128,1536,0.01275199982855055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,1536,0.006625777731339137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,1536,0.1421697802013821
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,128,768,0.0035146665241983202
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,128,1024,0.01240799989965227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,1024,0.006269333263238271
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,1024,0.1463279989030626
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,128,512,0.0032319999817344877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,128,4096,0.00666133314371109
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,768,0.006223111102978389
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,128,512,0.012108444339699216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,512,0.1463057729932997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,128,256,0.0029093333416514923
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,512,0.005906666732496685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,128,256,0.011128000087208219
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,256,0.14393866062164307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,128,128,0.0028826666788922418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,256,0.005592888842026393
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,128,128,0.011728888584507836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,128,0.1414462195502387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,128,64,0.002495999965402815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,128,2048,0.013076444466908773
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,128,32,0.0028159999185138275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,128,64,0.01071911139620675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,64,65536,0.013767999907334646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,128,32,0.011416888899273343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,64,16384,0.008010666403505538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,64,65536,0.0821582211388482
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,64,12288,0.007638221813572778
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,64,16384,0.027116444375779893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,768,0.1409484412935045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,64,10240,0.007090667055712805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,64,8192,0.006637333167923822
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,96,128,4096,0.13858221636878118
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,64,12288,0.02204799983236525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,64,7168,0.006575999988449945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,64,6144,0.006268444574541516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,64,8192,0.018533334136009216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,64,5120,0.006229333165619109
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,64,7168,0.01719377769364251
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,64,4096,0.005625777774386936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,64,6144,0.01629599928855896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,64,3584,0.006451555424266391
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,64,5120,0.015467555986510383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,64,3072,0.005959110955397288
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,64,4096,0.014118222726715935
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,64,2560,0.005992888990375731
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,2560,0.007292444507280986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,96,128,128,0.0059004442559348206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,64,2048,0.006218666831652324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,64,3072,0.013739555246300168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,64,1536,0.004514666481150521
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,64,1024,0.0038817777401871155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,64,2048,0.013139555851618448
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,64,768,0.0035386664999855887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,64,1536,0.01240799989965227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,64,512,0.0032337777730491427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,64,1024,0.01239733315176434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,64,256,0.0028533333291610083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,64,768,0.012309333516491784
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,64,128,0.002491555487116178
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,128,768,0.012101333174440594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,64,512,0.011713777979214987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,64,64,0.002513777878549364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,64,32,0.0028159999185138275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,64,128,0.011445333560307821
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,64,64,0.011422221859296164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,32,65536,0.01241422196229299
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,64,10240,0.02055288851261139
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,32,16384,0.007288888924651676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,64,32,0.011367999845080905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,32,12288,0.007305777735180325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,32,10240,0.007294221884674496
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,32,65536,0.08120622237523396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,32,16384,0.027373333772023518
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,32,8192,0.006996444529957241
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,32,12288,0.022616000639067754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,32,7168,0.006659555352396435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,32,6144,0.006463999963468975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,32,10240,0.020609777834680345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,32,5120,0.006264888991912206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,32,8192,0.018570666511853535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,32,7168,0.01756000022093455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,32,4096,0.005929777605666055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,64,3584,0.014178666803571912
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,32,6144,0.01649688846535153
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,32,3584,0.006620444357395172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,32,3072,0.006225777582989798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,32,5120,0.01555288831392924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,32,2560,0.006298666612969504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,64,2560,0.013105777402718862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,32,3584,0.014270222849316068
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,32,3072,0.013758222262064615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,32,1536,0.004497777670621872
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,32,2560,0.013764444324705334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,32,2048,0.012738666600651212
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,32,768,0.0035439998739295532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,32,1536,0.012711111042234631
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,32,512,0.00319377767542998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,64,256,0.011744889120260874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,32,768,0.011682666838169098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,32,256,0.0028151110228565005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,32,512,0.011755555868148804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,32,128,0.0028488888508743713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,32,64,0.0027991111079851785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,32,32,0.0028035555862718155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,32,256,0.011062222222487131
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,32,128,0.011713777979214987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,32,64,0.011403555671374003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,32,32,0.011341333389282227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,65536,16384,0.21502133210500082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,65536,16384,0.3621644443935818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,32,4096,0.014460444450378418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,32,2048,0.005943111247486538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,16384,0.14018221696217856
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,96,32,1024,0.0035386664999855887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,65536,12288,0.1593653361002604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,96,32,1024,0.012412444584899478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,65536,10240,0.21389599641164145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,65536,10240,0.13633422056833902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,12288,0.10540355576409234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,65536,8192,0.17794756094614664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,10240,0.08494044674767388
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,8192,0.06934489144219293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,16384,0.31173067622714573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,12288,0.24834489822387695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,65536,7168,0.15145156118604872
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,10240,0.21304710706075033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,8192,0.18120977613661024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,65536,6144,0.13142044014400905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,7168,0.06714222166273329
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,65536,6144,0.08753955364227295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,65536,8192,0.1162524488237169
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,6144,0.0539555549621582
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,65536,5120,0.11398043897416855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,7168,0.16680088308122423
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,65536,5120,0.07480800151824951
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,65536,4096,0.09174844291475083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,6144,0.15018755859798855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,5120,0.04727022184265984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,65536,4096,0.06299200322892931
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,65536,12288,0.2743519941965739
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,65536,3584,0.07997422085867988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,4096,0.03973333372010125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,5120,0.14344444539811876
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,65536,3584,0.05706666575537788
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,65536,3072,0.06950577762391832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,65536,7168,0.0994559990035163
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,3584,0.03527377711402045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,4096,0.14258044295840794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,65536,2560,0.058378669950697154
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,3584,0.14328444004058838
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,65536,2560,0.0454364452097151
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,65536,2048,0.047544889979892306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,3072,0.14201777511172825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,2560,0.14190577136145696
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,65536,2048,0.03992799917856852
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,65536,1536,0.03675288955370585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,2048,0.02440622283352746
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,2048,0.14267288313971624
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,65536,1536,0.04300622145334879
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,1536,0.1464080015818278
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,65536,1024,0.030702223380406696
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,65536,768,0.02126666737927331
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,1024,0.015473778049151102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,1024,0.14934933185577393
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,768,0.013784888717863293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,65536,768,0.0258577780591117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,768,0.1471395492553711
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,65536,512,0.01460444430510203
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,65536,3072,0.051023112403021924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,65536,256,0.009291555318567488
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,3072,0.0314044455687205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,512,0.012522666818565793
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,65536,512,0.024030221833123103
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,512,0.145687116516961
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,65536,128,0.007594666547245449
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,65536,256,0.018908444378111098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,2560,0.02718044486310747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,256,0.01109155515829722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,256,0.14420178201463488
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,65536,64,0.006711111300521427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,65536,128,0.017861333158281114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,65536,32,0.006937777830494775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,65536,128,0.13440355989668104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,128,0.011010666688283285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,65536,64,0.01719199948840671
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,65536,32,0.022284444835450914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,16384,16384,0.08250844478607178
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,65536,1024,0.027753776974148218
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,16384,65536,0.32317956288655597
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,16384,65536,0.26139020919799805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,16384,16384,0.0746062199274699
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,16384,12288,0.06422666708628337
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,16384,0.04346488912900289
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,16384,12288,0.056629333231184215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,65536,0.15119466516706678
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,16384,10240,0.05420355664359199
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,16384,0.14767466651068792
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,12288,0.1410026682747735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,16384,10240,0.04951022068659464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,16384,8192,0.045531554354561694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,10240,0.029899554120169744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,16384,7168,0.039375109805001154
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,16384,8192,0.04355822337998284
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,10240,0.1431893375184801
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,8192,0.024451555477248296
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,16384,7168,0.03846933444341024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,8192,0.1446062193976508
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,16384,6144,0.03456977672047085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,7168,0.023846222294701472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,7168,0.1423057715098063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,16384,5120,0.02958844436539544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,16384,6144,0.034671111239327326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,6144,0.019834667444229126
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,6144,0.1459911134507921
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,16384,5120,0.03120622369978163
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,16384,4096,0.02549244463443756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,5120,0.14050666491190592
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,16384,4096,0.027444443768925134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,65536,0.2806382179260254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,4096,0.015392000476519266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,16384,3584,0.02180800007449256
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,4096,0.14382755756378174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,16384,3072,0.01955111159218682
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,12288,0.033717334270477295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,3584,0.13943644364674887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,16384,3584,0.0251857770813836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,16384,2560,0.01622222198380364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,16384,3072,0.02346399923165639
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,3072,0.013089777694808112
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,3072,0.1448942157957289
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,16384,2048,0.013429332938459186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,16384,2560,0.021291555629836187
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,2560,0.012147555748621622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,2560,0.1439946624967787
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,16384,1536,0.010874666273593903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,16384,2048,0.01982666717635261
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,2048,0.011351111034552256
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,16384,1536,0.019837333096398246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,16384,1024,0.008830221990744272
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,65536,1536,0.020248888267411124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,1536,0.14727911684248182
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,5120,0.01757777730623881
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,16384,768,0.007260444263617198
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,16384,1024,0.01686844395266639
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,1024,0.14095555411444768
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,1024,0.008277333445019191
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,16384,512,0.006103111224042044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,16384,768,0.015372445185979208
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,768,0.1445066663953993
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,16384,512,0.014818666709793938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,16384,256,0.0038942222793896994
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,512,0.007274666594134436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,3584,0.014529777897728814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,16384,256,0.013068444199032254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,256,0.14466133382585314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,256,0.006620444357395172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,16384,128,0.013091555900043912
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,128,0.13180443975660536
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,16384,64,0.003516444315512975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,2048,0.14232799741956922
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,128,0.006823110911581252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,16384,32,0.0034764446318149567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,16384,64,0.013105777402718862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,16384,32,0.01479466590616438
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,1536,0.009680888719028896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,12288,65536,0.2551928891075982
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,12288,65536,0.20565599865383574
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,12288,16384,0.07117155525419447
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,16384,768,0.007620444728268518
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,16384,512,0.1474213335249159
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,12288,16384,0.061921775341033936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,16384,0.04090133309364319
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,12288,12288,0.055200000603993736
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,16384,128,0.003510222252872255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,12288,12288,0.047304888566335045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,16384,0.1431528859668308
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,12288,10240,0.04575466778543261
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,12288,0.031574222776624895
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,12288,8192,0.03911644551489089
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,12288,0.14182844426896837
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,12288,10240,0.04142489035924276
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,10240,0.02681155502796173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,10240,0.14414578013949925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,12288,7168,0.03379199902216593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,12288,8192,0.036588443650139704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,8192,0.022110222114457026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,12288,6144,0.029399110211266413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,8192,0.13990399572584364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,12288,7168,0.03286222285694546
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,7168,0.020611556039916146
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,7168,0.14184266991085476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,12288,6144,0.030064887470669214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,6144,0.018238221605618794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,12288,4096,0.021923555268181696
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,6144,0.14730932977464464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,12288,5120,0.02562933332390255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,5120,0.0165928883685006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,5120,0.141412443584866
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,12288,3584,0.019015999303923715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,12288,4096,0.02330933345688714
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,4096,0.014702222413486905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,4096,0.14750311109754774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,65536,0.1341164509455363
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,12288,3072,0.016275554895401
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,3584,0.013527111046844058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,3584,0.1409333282046848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,12288,3584,0.021890666749742296
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,12288,2560,0.014070222775141398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,3072,0.012978666358523898
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,12288,2048,0.012099555797047086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,12288,2560,0.01920977731545766
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,2560,0.01146399974822998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,2560,0.14237956206003824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,12288,2048,0.018090666996108163
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,2048,0.13969866434733072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,2048,0.010058666268984476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,12288,1536,0.009957333405812582
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,12288,1536,0.019723556107944913
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,1536,0.00850399997499254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,12288,1024,0.0084915558497111
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,12288,5120,0.025390222668647766
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,65536,0.2456515630086263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,12288,1024,0.015383111106024848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,1024,0.14236177338494196
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,1024,0.007822222179836696
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,12288,768,0.014133334159851074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,12288,512,0.00499199993080563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,768,0.13959466086493597
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,12288,512,0.014112000664075216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,12288,3072,0.020503110355801053
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,3072,0.13964800039927164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,12288,256,0.0038817777401871155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,512,0.006904888898134232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,12288,256,0.012410666379663678
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,256,0.006587555425034628
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,256,0.14762755235036215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,12288,128,0.012703110774358114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,128,0.13755022154914007
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,12288,64,0.0032417778339650896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,1536,0.14310844739278158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,128,0.006587555425034628
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,12288,32,0.0034737777378824023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,12288,64,0.012068444656001197
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,12288,32,0.013766222529941134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,12288,768,0.0069324444565508105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,10240,65536,0.21783199575212267
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,12288,768,0.007260444263617198
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,10240,65536,0.19955200619167754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,10240,16384,0.055953780810038246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,12288,512,0.1386000050438775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,65536,0.1306355529361301
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,16384,0.036892443895339966
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,10240,12288,0.04422577884462145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,16384,0.14832622475094265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,10240,12288,0.04565066761440701
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,12288,128,0.0035297779573334586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,12288,0.02903911140229967
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,10240,10240,0.03870844509866502
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,12288,0.14177244239383274
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,10240,10240,0.040375110175874494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,10240,8192,0.03183199961980184
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,10240,0.025011556016074285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,10240,0.14099110497368708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,10240,7168,0.02977777851952447
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,10240,8192,0.035995556248558894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,8192,0.021294222937689886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,65536,0.23643732070922852
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,10240,7168,0.0319164428446028
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,7168,0.019871110717455547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,7168,0.1383555597729153
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,10240,5120,0.02141511109140184
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,10240,6144,0.028816001282797918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,6144,0.01760088900725047
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,6144,0.14035022258758545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,10240,4096,0.01896711190541585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,10240,5120,0.02502311103873783
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,5120,0.015987555185953777
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,5120,0.1412995523876614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,10240,3584,0.016514667206340365
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,4096,0.014087999860445658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,10240,3584,0.02162488963868883
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,3584,0.013242666920026144
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,3584,0.1425066656536526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,10240,3072,0.020624889267815482
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,3072,0.14206133948432073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,10240,2560,0.012832889126406776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,3072,0.012485333614879184
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,8192,0.145050671365526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,10240,2560,0.01903288894229465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,10240,2048,0.011257777611414591
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,2560,0.14260355631510416
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,2560,0.011032889286677042
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,10240,2048,0.018039110634062026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,2048,0.14483733971913657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,10240,1536,0.00960799968904919
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,2048,0.009673777553770278
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,10240,16384,0.0595911078982883
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,10240,1536,0.018515555395020377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,1536,0.14010666476355657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,1536,0.008515555825498369
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,10240,4096,0.022758222288555566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,10240,768,0.006259555617968242
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,10240,1024,0.015048000547620984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,1024,0.007676444119877285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,1024,0.13909067047966853
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,10240,3072,0.014371555712487964
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,10240,512,0.0052639999323421055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,10240,768,0.014087999860445658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,768,0.007425777614116669
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,768,0.14332977930704752
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,10240,256,0.004224888980388641
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,10240,512,0.013436444103717804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,512,0.14159644974602595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,512,0.0069306666652361555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,10240,128,0.0036480000449551474
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,10240,256,0.01255022237698237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,256,0.14475733704037136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,256,0.006566222343179915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,10240,64,0.0035075553589397003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,10240,6144,0.024293333292007446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,128,0.13334843847486708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,10240,32,0.003491555651028951
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,10240,128,0.006599111275540457
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,10240,64,0.012768000364303589
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,10240,32,0.01348000019788742
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,8192,65536,0.17312888304392496
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,8192,65536,0.15754666593339708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,10240,1024,0.007989333735571967
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,10240,4096,0.1394231054517958
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,65536,0.07816888888676961
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,8192,16384,0.049138665199279785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,8192,12288,0.037652442852656044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,16384,0.02478755513827006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,8192,12288,0.03762933280732896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,8192,10240,0.031176888280444678
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,16384,0.14482133918338352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,12288,0.1419608857896593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,8192,10240,0.033599111768934466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,65536,0.2199537754058838
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,10240,0.01722311145729489
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,8192,8192,0.027072888281610277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,10240,128,0.012086222569147745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,8192,7168,0.0233262214395735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,8192,0.014783999986118741
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,8192,8192,0.029842668109469946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,8192,0.1422462198469374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,8192,7168,0.02647288971477085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,8192,6144,0.02032888929049174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,7168,0.014013333453072442
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,8192,16384,0.04714222086800469
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,7168,0.145961774720086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,8192,6144,0.023895111348893907
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,8192,5120,0.017512000269360013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,6144,0.14258489343855116
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,8192,4096,0.01574577722284529
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,8192,5120,0.021828444467650518
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,5120,0.14195644855499268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,5120,0.011728000309732227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,8192,4096,0.020265777905782063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,8192,3584,0.013422222601042854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,4096,0.010696888797812991
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,8192,3072,0.012059555285506778
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,8192,3584,0.019508444600635104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,3584,0.14295021692911783
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,3584,0.009711999860074785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,12288,0.019318222999572754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,8192,2560,0.01037688884470198
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,8192,3072,0.01831733352608151
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,3072,0.009025777379671732
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,3072,0.1426195568508572
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,8192,2560,0.01715555621518029
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,10240,0.14490044116973877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,8192,2048,0.009291555318567488
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,8192,1536,0.007989333735571967
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,2560,0.14816978242662218
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,8192,2048,0.016806221670574613
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,2048,0.007648888561460707
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,8192,1536,0.016757332616382174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,8192,1024,0.006151111175616582
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,1536,0.14737244447072348
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,1536,0.00721066693464915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,8192,768,0.004888888862397936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,8192,1024,0.014096000128322177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,1024,0.006601777755551868
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,1024,0.1450222200817532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,8192,512,0.0041955556306574075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,8192,768,0.013992889059914483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,6144,0.012870222330093384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,768,0.14256800545586482
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,8192,256,0.0034871109657817413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,8192,512,0.013061333033773633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,512,0.006281777802440856
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,512,0.14230399661593968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,8192,128,0.0032337777730491427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,8192,256,0.012092444631788464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,4096,0.14459289444817439
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,256,0.006055999961164262
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,256,0.1452702283859253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,8192,64,0.002833777831660377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,8192,128,0.011758222348160215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,128,0.005953777581453323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,8192,32,0.0031964443624019623
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,128,0.13393155733744302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,8192,64,0.012064889073371887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,8192,32,0.012440889245933957
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,7168,65536,0.1534951130549113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,7168,16384,0.046203556987974376
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,7168,65536,0.15377067195044622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,8192,2048,0.14187199539608425
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,65536,0.07539200120502047
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,7168,12288,0.03418488966094123
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,16384,0.023927999867333308
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,16384,0.14175289207034641
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,7168,12288,0.0365573333369361
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,7168,10240,0.029145777225494385
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,12288,0.14000711176130506
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,12288,0.01864177816443973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,768,0.0063768890168931745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,7168,8192,0.02437955637772878
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,65536,0.21675911214616564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,7168,10240,0.0325857765144772
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,10240,0.1399422221713596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,7168,7168,0.02169599963559045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,7168,8192,0.028794666131337483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,8192,0.014489778214030795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,8192,0.14202488793267146
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,7168,7168,0.025624000363879736
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,7168,0.013293333351612091
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,7168,0.14658310678270128
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,8192,2560,0.008000888758235509
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,6144,0.01239911135700014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,7168,5120,0.01627288924323188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,6144,0.14524355199601915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,7168,16384,0.04786044359207153
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,7168,5120,0.02121155626244015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,7168,4096,0.014731556177139282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,7168,4096,0.019574221637513902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,4096,0.1450222200817532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,7168,3584,0.012612444659074148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,4096,0.010004444254769219
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,10240,0.016536000702116225
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,7168,3584,0.018924444913864136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,7168,3072,0.011382222175598145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,3584,0.14856533209482828
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,3584,0.008851555486520132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,7168,3072,0.018213333355055917
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,3072,0.008295999632941352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,7168,2560,0.010110222631030613
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,3072,0.146598219871521
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,7168,6144,0.023472888602150813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,2560,0.007979555262459649
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,2560,0.14755554993947348
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,7168,2048,0.009010666774378883
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,7168,2048,0.015777778294351365
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,2048,0.1458302206463284
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,5120,0.011525332927703857
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,7168,1536,0.007631111476156447
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,5120,0.14056532912784153
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,7168,1024,0.005816000203291575
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,7168,1536,0.015847111741701763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,1536,0.00701422244310379
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,1536,0.14431199762556288
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,7168,768,0.004524444540341695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,7168,1024,0.013957333233621387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,1024,0.006596444381607904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,1024,0.1450773345099555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,7168,512,0.0038702223036024305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,7168,768,0.013120888835854001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,768,0.006299555715587404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,768,0.13955733511182997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,7168,256,0.003547555456558863
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,7168,512,0.013094222380055321
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,512,0.006272000157170826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,512,0.1449084414376153
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,7168,256,0.012072000238630505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,7168,128,0.0031600000543726813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,7168,6144,0.019133332702848647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,256,0.005930666708283954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,7168,64,0.0028400001012616684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,7168,128,0.011733333269755045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,128,0.1363742219077216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,7168,32,0.0030453333424197305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,7168,2560,0.017151999804708693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,7168,64,0.011384888655609555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,7168,32,0.013117333253224691
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,6144,65536,0.1275964445537991
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,2048,0.0076044441925154785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,65536,0.07348889112472534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,6144,16384,0.04640355706214905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,6144,12288,0.033357332150141396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,16384,0.14093689123789468
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,16384,0.022487999664412603
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,6144,12288,0.03597866826587253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,12288,0.018200889229774475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,6144,10240,0.029334220621320937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,65536,0.21504888269636366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,12288,0.14012800322638616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,6144,8192,0.021574222379260596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,6144,10240,0.030984888474146526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,10240,0.016185777054892648
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,10240,0.14624444643656412
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,7168,256,0.14430133501688638
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,6144,7168,0.019848888119061787
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,8192,0.014147554834683737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,7168,128,0.0059057776298787855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,6144,7168,0.025107555919223364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,7168,0.1453582180870904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,7168,0.01309333327743742
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,6144,6144,0.023299554983774822
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,6144,16384,0.04016800059212579
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,6144,5120,0.014807111687130399
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,6144,0.012096000214417776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,6144,0.14489689138200548
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,6144,5120,0.02128000060717265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,5120,0.010688888529936472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,6144,4096,0.013072888884279462
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,6144,4096,0.019539554913838703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,4096,0.1402933332655165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,6144,3584,0.011937778029176923
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,4096,0.00924177798959944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,6144,3584,0.018836443622907
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,6144,8192,0.026995556222067937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,3584,0.14482577641805014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,6144,3072,0.010997333460383944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,3584,0.009036444127559662
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,8192,0.14655289385053846
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,6144,2560,0.010011555420027839
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,6144,6144,0.018039999736679923
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,3072,0.008293333152929941
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,3072,0.14226667086283365
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,6144,2048,0.008443555898136562
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,6144,65536,0.14827466011047363
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,6144,2048,0.015142222245534262
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,2048,0.14075377252366808
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,5120,0.1458524465560913
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,6144,1536,0.007280888656775157
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,6144,1536,0.01514844430817498
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,1536,0.14082133769989014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,6144,1024,0.00556355549229516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,1536,0.006937777830494775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,6144,1024,0.013781333135233985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,1024,0.1455004480150011
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,6144,768,0.004551110996140374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,1024,0.006559999866618051
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,6144,3072,0.017605332864655387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,6144,768,0.013429332938459186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,768,0.14145689540439182
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,768,0.006256000035338932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,6144,2560,0.016344888342751395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,2560,0.007879111501905654
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,6144,512,0.01278222186697854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,6144,256,0.0034826666944556763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,512,0.0059279998143513995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,512,0.14551644855075413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,6144,256,0.01204533295498954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,6144,128,0.0031519999934567344
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,256,0.1409848928451538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,256,0.006093333164850871
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,6144,64,0.0029120000286234748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,6144,128,0.011742221812407175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,6144,32,0.0031475555151700974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,6144,64,0.011375111010339526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,6144,32,0.012314666476514606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,5120,65536,0.11795377731323242
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,5120,65536,0.14448622862497965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,5120,16384,0.03444177905718485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,65536,0.0709644423590766
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,2560,0.14232977231343588
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,5120,16384,0.04494577646255493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,5120,12288,0.029843555556403265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,16384,0.02109066645304362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,16384,0.14098666773902044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,2048,0.007316444483068254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,5120,12288,0.034632888105180525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,5120,10240,0.02502222193611993
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,12288,0.01737155516942342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,6144,128,0.005897777775923411
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,12288,0.14239911238352457
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,5120,8192,0.022256000174416438
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,5120,10240,0.030304001437293157
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,10240,0.015767110718621146
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,10240,0.14124800099266901
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,5120,7168,0.019333332777023315
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,8192,0.013738666971524557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,8192,0.13958577315012613
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,5120,6144,0.014799111419253878
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,7168,0.14069155851999918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,5120,6144,0.022600889205932617
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,6144,0.14019821749793157
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,6144,512,0.0038613333470291565
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,5120,5120,0.012904889053768583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,6144,0.011711999773979187
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,5120,5120,0.021037333541446265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,5120,0.010501333408885531
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,5120,4096,0.01126488877667321
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,5120,0.1387057834201389
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,65536,0.2142773336834378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,5120,4096,0.01923822197649214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,4096,0.009292444421185387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,4096,0.14682755205366346
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,6144,128,0.13569600052303737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,5120,3584,0.018541332748201158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,5120,3072,0.010334221853150262
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,5120,3072,0.017148445049921673
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,5120,8192,0.027048889133665297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,3072,0.14324533939361572
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,5120,7168,0.02444711161984338
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,3072,0.007986666427718269
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,7168,0.012655110822783576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,5120,2560,0.016024000114864774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,5120,2048,0.008641777767075432
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,2560,0.007647111184067196
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,2560,0.14225688245561388
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,5120,2048,0.0147733340660731
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,5120,1536,0.006918222126033571
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,2048,0.007256888680987888
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,5120,1536,0.01442399952146742
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,1536,0.1416862275865343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,5120,1024,0.00592533333433999
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,1536,0.006907555378145642
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,5120,3584,0.01035288886891471
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,5120,1024,0.013627555635240344
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,5120,768,0.004858666823969947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,1024,0.1456800036960178
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,3584,0.008658666577604082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,3584,0.14067999521891275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,5120,512,0.0041626666982968645
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,5120,768,0.013056889176368713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,768,0.006274666637182236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,768,0.14645511574215359
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,5120,2560,0.008996444443861643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,5120,512,0.01274577776590983
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,512,0.006106666806671355
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,512,0.14015911685095891
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,5120,128,0.003177777760558658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,5120,256,0.012053333222866058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,256,0.005930666708283954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,5120,128,0.012072000238630505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,128,0.133114669058058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,2048,0.14105244477589926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,128,0.006074666563007567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,5120,64,0.0031511110977994073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,5120,32,0.0032142222755485107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,5120,64,0.012153777811262341
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,5120,32,0.012811555630630918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,4096,65536,0.08709155850940281
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,4096,16384,0.03019022279315525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,4096,65536,0.10361244281133015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,65536,0.06872089041603936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,5120,1024,0.006596444381607904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,4096,16384,0.03360622127850851
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,4096,12288,0.024480000138282776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,16384,0.14166133933597141
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,16384,0.020787555310461257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,4096,12288,0.0262737770875295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,4096,10240,0.021505778034528095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,5120,256,0.0037502224246660867
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,12288,0.016886222693655226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,12288,0.1472595532735189
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,4096,8192,0.015761777758598328
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,4096,10240,0.023644444015291
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,10240,0.015169777803950839
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,10240,0.14605511559380427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,4096,7168,0.014277334014574686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,4096,8192,0.021401777863502502
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,8192,0.14194044801923963
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,8192,0.01309866706530253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,4096,6144,0.012753778033786349
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,4096,7168,0.01984177695380317
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,7168,0.1441804435518053
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,7168,0.012063110868136087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,4096,5120,0.011513777905040316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,4096,6144,0.018896889355447557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,6144,0.01127911110719045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,6144,0.14092889097001818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,4096,4096,0.0107022225856781
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,4096,5120,0.01724088854259915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,5120,0.0099973330895106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,5120,0.14410221576690674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,4096,3584,0.009287111461162567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,4096,4096,0.016494222813182406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,4096,0.00869599978129069
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,4096,3584,0.01553600033124288
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,4096,3072,0.008300444318188561
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,3584,0.1422604454888238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,4096,3072,0.015118221441904703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,3072,0.007923555870850882
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,3072,0.14905067284901938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,4096,2560,0.007728888756699032
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,65536,0.21177777979109022
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,4096,2048,0.006511111226346757
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,4096,2560,0.0143857780430052
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,2560,0.0075786668393347
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,2560,0.14726044072045222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,4096,1536,0.005592000153329637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,4096,2048,0.013461333182122974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,2048,0.1388817760679457
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,5120,256,0.1398515568839179
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,4096,1024,0.00423200014564726
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,4096,1536,0.013787555197874704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,1536,0.0069146665434042616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,1536,0.14499821927812365
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,4096,768,0.003847111016511917
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,4096,1024,0.012727999852763282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,1024,0.1401475535498725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,1024,0.006616000086069107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,4096,512,0.003527111063400904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,4096,768,0.012404444317022959
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,768,0.006237333433495627
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,768,0.14429599708980984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,4096,256,0.0031991110493739447
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,4096,512,0.012050666742854647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,512,0.005958222266700532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,512,0.13929777675204807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,4096,256,0.011762667033407422
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,256,0.005879999862776862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,256,0.1416355503929986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,4096,128,0.011728000309732227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,128,0.005936000082227919
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,4096,64,0.002793777734041214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,128,0.13791555828518337
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,4096,32,0.003154666680428717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,4096,64,0.011519999967681037
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3584,65536,0.07701155874464247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,4096,32,0.011744000017642975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3584,65536,0.10166044367684259
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,2048,0.007152888509962294
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,65536,0.06775466601053874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3584,16384,0.026918222506841023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3584,16384,0.03235377868016561
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,16384,0.020257777637905545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3584,12288,0.021007999777793884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,65536,0.21132889058854845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3584,12288,0.024813334147135418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3584,10240,0.018127111925019156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,12288,0.016743110285864938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,4096,128,0.002842666581273079
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,4096,4096,0.14496889379289415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3584,8192,0.017214222086800467
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,10240,0.015090665883488126
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,4096,3584,0.008285333712895712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,10240,0.14369600349002415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3584,7168,0.015785778562227886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3584,8192,0.021241777473025854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,8192,0.13921688662634954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3584,7168,0.019904888338512845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3584,6144,0.014868444866604276
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,7168,0.011799111134476133
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,7168,0.14518045054541692
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3584,6144,0.018937778141763475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,6144,0.011376889215575324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,6144,0.14084445105658636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3584,5120,0.017698667115635343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,5120,0.13946844471825495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,16384,0.14514488644070095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3584,4096,0.010062221851613786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,5120,0.010199999643696679
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3584,4096,0.015585778488053216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,12288,0.1410799953672621
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3584,3584,0.008988444175985124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,4096,0.13932355244954428
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,4096,0.008671110702885522
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3584,3072,0.007961778177155389
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3584,3584,0.0148062225845125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3584,10240,0.02294755478700002
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,3584,0.00829777783817715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,3584,0.14537155628204346
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3584,2560,0.007328000333574083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3584,3072,0.014793778459231058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,3072,0.007923555870850882
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,3072,0.1406444443596734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3584,2048,0.006219555520349079
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,8192,0.012783110969596438
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,2560,0.007746666669845581
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,2560,0.1397520038816664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3584,1536,0.005579555614127054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3584,2048,0.01346577786737018
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,2048,0.007330666813585494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,2048,0.13945066928863525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3584,1024,0.004223111189073986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3584,1536,0.013449778159459433
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,1536,0.006981333096822103
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,1536,0.1395555602179633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3584,768,0.003863999827040566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3584,1024,0.012737777498033313
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3584,5120,0.010984000232484607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,1024,0.14170311556922063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3584,512,0.003516444315512975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3584,768,0.012415111064910889
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,768,0.006272000157170826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,768,0.13883288701375326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3584,256,0.0031946665710873077
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3584,512,0.012256888879670037
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,512,0.005968889014588461
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,512,0.1394382185406155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3584,256,0.011437333292431302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,256,0.005942222144868638
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3584,128,0.0028239999794297745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,256,0.13956444793277317
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3584,64,0.0028453332682450614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3584,32,0.0028577778074476454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3584,128,0.011360000405046674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,128,0.0059057776298787855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,3584,128,0.13229778077867296
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3584,64,0.011754666765530905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3072,65536,0.06900799936718412
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3584,32,0.012220444778601328
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3072,16384,0.022308443983395893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3072,65536,0.09962577952278985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3072,16384,0.031086222993002996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,65536,0.06673333379957411
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,16384,0.020055999358495075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,16384,0.14591911103990343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3072,12288,0.019256888164414298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3072,10240,0.01626044511795044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3584,2560,0.013743110828929476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3072,12288,0.024695111645592585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,12288,0.1409013271331787
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,65536,0.21112710899776885
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3072,8192,0.014090667168299357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3072,10240,0.022456889351209004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,10240,0.01479822231663598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,10240,0.1442479954825507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3072,7168,0.013068444199032254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3072,8192,0.020634666085243225
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,8192,0.012267555627557965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,3584,1024,0.006270222365856171
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3072,7168,0.019066666563351948
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,7168,0.011425777441925473
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,7168,0.14046666357252333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3072,5120,0.010779555473062726
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3072,6144,0.01832799944612715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,6144,0.010258666343159145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,6144,0.14605955282847086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3072,4096,0.009355555805895064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3072,5120,0.01700444519519806
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,5120,0.009383111364311641
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,5120,0.14136711756388345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3072,3584,0.009277332988050249
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3072,4096,0.015814221567577787
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,4096,0.008468444148699442
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,4096,0.13917777273390028
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3072,3072,0.008208889100286696
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3072,3584,0.015150222513410779
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,3584,0.007988444632954067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,3584,0.14432444837358263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3072,2560,0.007282666862010956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3072,3072,0.014584889014561972
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,3072,0.13963022496965197
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,3072,0.007840889195601145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,12288,0.01649777756796943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3072,2048,0.006603555546866522
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3072,2560,0.013753777576817406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,2560,0.14040799935658774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,2560,0.007311999797821045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3072,1536,0.005199111170238919
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3072,2048,0.013437333206335703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,2048,0.007195555501514011
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,2048,0.14864533477359348
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3072,1024,0.004571555389298333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,8192,0.141047994295756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3072,1536,0.013247110777431063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,1536,0.0069671107663048645
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3072,768,0.003834666477309333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,1536,0.14081422487894693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3072,1024,0.012402666939629449
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,1024,0.0063386667105886675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,1024,0.14576266871558294
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3072,512,0.0035626664757728577
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3072,768,0.012431110772821637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,768,0.006214222146405115
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,768,0.1403262217839559
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3072,256,0.0031911111954185697
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,512,0.005919999960396025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,512,0.14466755919986302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3072,128,0.0028533333291610083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3072,256,0.011723555624485016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,256,0.14056266678704157
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3072,64,0.0028026666906144884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3072,128,0.011370666325092316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,128,0.005869333528810077
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,3072,128,0.1383999983469645
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3072,32,0.002941333378354708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3072,64,0.011730666789743634
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3072,32,0.01238666640387641
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2560,65536,0.09590755568610297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2560,16384,0.020557334025700886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,65536,0.06557777855131361
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2560,16384,0.02999733222855462
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,65536,0.2110817829767863
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,16384,0.14132977856530085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,3072,6144,0.012774222426944308
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,16384,0.019945777124828763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2560,12288,0.024332443873087566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,12288,0.015957333975368075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,12288,0.14286222722795275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,3072,512,0.012033777932325998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,3072,256,0.005933333188295364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2560,10240,0.022104889154434204
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2560,8192,0.013769778112570444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,10240,0.14078488614824083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2560,8192,0.02015199926164415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,8192,0.011697777443461947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,8192,0.14577066898345947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2560,65536,0.06040088997946846
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2560,6144,0.011424000064531961
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2560,7168,0.019038221902317472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,7168,0.01072088877360026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,7168,0.1450631088680691
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2560,5120,0.01036622209681405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2560,6144,0.017647110753589206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,6144,0.010035555395815108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,6144,0.14001866181691489
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2560,5120,0.016715556383132935
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2560,4096,0.00904977818330129
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,5120,0.14799733956654867
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2560,12288,0.01777333352300856
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2560,4096,0.01513777838812934
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,4096,0.1483128865559896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,4096,0.008344000412358178
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2560,10240,0.015579554769727917
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2560,3584,0.014466666513019137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,3584,0.14113333490159777
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,3584,0.008273777862389883
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,10240,0.014093332820468478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2560,3072,0.014161777165200977
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2560,2560,0.0069511110583941145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,3072,0.1406008932325575
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2560,7168,0.013103110922707452
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2560,2560,0.013779555757840475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,2560,0.14197688632541233
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2560,2048,0.0058808889653947614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2560,2048,0.013403555585278405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,2048,0.14074044757419163
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2560,1536,0.004844444493452708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,2048,0.007256000406212277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,5120,0.009697777529557547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2560,1536,0.013054221868515015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2560,1024,0.004516444272465176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2560,3584,0.0083137775460879
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,1536,0.14160888724856907
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2560,1024,0.01242844429281023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,1024,0.0063884444534778595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2560,768,0.0038417776425679526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,1024,0.1443271107143826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2560,3072,0.007781333393520779
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2560,512,0.0034755555291970572
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2560,768,0.012072888513406118
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,768,0.006037333359320958
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,768,0.14142489433288574
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2560,256,0.0031902222997612427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2560,512,0.012091555529170565
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,3072,0.00792266676823298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,512,0.005936000082227919
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,512,0.1414737833870782
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2560,128,0.003133333391613431
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2560,256,0.011692444483439127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,256,0.005879999862776862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2560,64,0.00281866660548581
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,2560,0.007599111232492659
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2560,32,0.0028275555620590844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,128,0.005585777676767773
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,128,0.1357075505786472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2560,64,0.011360888679822287
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2048,65536,0.04997244477272034
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2560,32,0.01145333300034205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2048,16384,0.01848266687658098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2048,65536,0.091648002465566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,65536,0.06416800287034777
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,16384,0.14301155673133
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,16384,0.019350222415394254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,65536,0.20902933014763725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,2560,1536,0.0069511110583941145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2048,10240,0.013750221994188098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2048,12288,0.02416177756256527
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,12288,0.015473778049151102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,12288,0.14099200566609701
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2048,8192,0.012725333372751871
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2048,10240,0.021919111410776775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,10240,0.013430222041077085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2048,8192,0.020096888144810993
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2048,7168,0.011677333050303988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,8192,0.011388444238238864
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2048,7168,0.018408889571825664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,2560,256,0.1470995611614651
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,7168,0.14618578222062853
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2560,128,0.011666666302416058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2048,6144,0.01125511113140318
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,7168,0.01089599976936976
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2048,5120,0.010079999764760336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,6144,0.010092444717884064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,6144,0.14091911580827501
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2048,5120,0.015838222371207345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2048,16384,0.029473778274324205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,5120,0.009341333475377824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,5120,0.1476044389936659
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2048,4096,0.01148888882663515
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2048,12288,0.014803555276658801
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2048,3584,0.010309333602587381
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,4096,0.008358221915033128
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,4096,0.14008355140686035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2048,3584,0.014464888307783337
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2048,3072,0.00795199970404307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,3584,0.13972444004482693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,3584,0.007981333467695449
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2048,2560,0.007617777420414819
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,3072,0.007642666498819987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2048,3072,0.014073777529928418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,3072,0.1398435566160414
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2048,2048,0.005543111099137201
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2048,2560,0.013454222016864352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,2560,0.007327111230956183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,8192,0.14213511678907606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,2560,0.14279733763800725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2048,1536,0.004804444395833545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2048,2048,0.013402666482660504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,2048,0.007095110913117726
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2048,1536,0.012780444489585029
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2048,1024,0.0042257776690853965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,1536,0.14121422502729628
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,1536,0.006815111057625876
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2048,6144,0.01719466679626041
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2048,768,0.003863111138343811
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2048,1024,0.012423999607563019
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,1024,0.1396204498079088
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2048,768,0.01240711079703437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,768,0.006218666831652324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,768,0.14101600646972656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2048,4096,0.01478488908873664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2048,512,0.012063110868136087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,512,0.14090577761332193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2048,256,0.0031848889258172777
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,512,0.005948444621430502
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,10240,0.14307822121514216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,256,0.005615111026499007
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,256,0.14178489314185247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2048,128,0.0028533333291610083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2048,128,0.011409777734014722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,128,0.13142755296495226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,2048,2048,0.147816883193122
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2048,32,0.0031519999934567344
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2048,64,0.011757333245542316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1536,65536,0.039093332158194646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,1024,0.006263999889294307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1536,65536,0.09004710780249701
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2048,512,0.003510222252872255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1536,16384,0.01756444407833947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,65536,0.2093493276172214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,65536,0.06259822183185153
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1536,16384,0.02918755677011278
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1536,12288,0.015474667151769003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,16384,0.018212444252438016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,16384,0.1413537793689304
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2048,256,0.011713777979214987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1536,10240,0.014020444618331062
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1536,12288,0.0236817780468199
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,12288,0.14109867148929173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,12288,0.015136000182893542
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,2048,128,0.005628444254398346
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1536,8192,0.012488889197508493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,2048,64,0.0028115556471877625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1536,10240,0.02161066730817159
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,10240,0.14631733629438612
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1536,7168,0.011734222372372946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,10240,0.013040000365840064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1536,8192,0.0192577772670322
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,8192,0.011368888947698804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,8192,0.1398871077431573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1536,6144,0.011365333365069495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1536,7168,0.01789155602455139
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,7168,0.010672888822025724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,7168,0.14219289355807835
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,2048,32,0.010808000134097205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1536,5120,0.009637333452701569
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1536,4096,0.008289777570300633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1536,6144,0.016501333978441026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,6144,0.009999111294746399
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,6144,0.14225867059495714
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1536,5120,0.01571822166442871
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,5120,0.009325332939624786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,5120,0.14262310663859049
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1536,4096,0.01481599940194024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,4096,0.00833511104186376
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1536,3072,0.0069422221018208405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,4096,0.14456621805826822
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1536,3584,0.01443733274936676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,3584,0.007981333467695449
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,3584,0.1490444474750095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1536,2560,0.006258666515350342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1536,3072,0.014102222190962898
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,3072,0.007712889048788283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,3072,0.14606044027540419
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1536,2048,0.005208888815508948
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,2560,0.007438222567240397
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,2560,0.14692533016204834
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1536,2048,0.01311555587583118
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,2048,0.006963555183675554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,2048,0.14591644869910345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1536,1024,0.0041546668443414895
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1536,1536,0.012755555411179861
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,1536,0.006611555400821898
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,1536,0.14587555991278753
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1536,1024,0.01237511138121287
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,1024,0.14539378219180638
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1536,768,0.012088889049159156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1536,512,0.0034942221310403612
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,768,0.006107555495368109
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,768,0.13993067211574978
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1536,512,0.012055111428101858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,512,0.0058906666106647914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1536,256,0.0031644445326593188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,512,0.14406666490766737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1536,3584,0.007582222421964009
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1536,256,0.011695110963450538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1536,128,0.0028568889117903183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,256,0.1458711094326443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1536,128,0.011396444506115384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,1536,128,0.1364977757136027
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1536,2560,0.013753777576817406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1536,64,0.002854222224818336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1536,1536,0.004902222090297275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1536,32,0.0028488888508743713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,128,0.005711110929648082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1024,65536,0.02749600013097127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1536,64,0.011408888631396823
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1536,32,0.012147555748621622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1024,16384,0.012857777376969656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1536,768,0.0038293335172865125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1024,65536,0.08698666758007473
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,1024,0.006310222049554189
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,65536,0.058712886439429395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1024,12288,0.010595555934641095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,65536,0.20882044898139107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1024,16384,0.02865866820017497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,16384,0.016905777984195285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,16384,0.143939561314053
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1024,10240,0.011680000358157687
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1024,12288,0.023131555981106226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,12288,0.014097778333557976
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1024,8192,0.01331555595000585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1024,10240,0.020233778489960563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,10240,0.0127697777416971
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,10240,0.14351644780900744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1024,7168,0.011723555624485016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1024,8192,0.018210666047202218
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,8192,0.1404551135169135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,8192,0.011352000137170156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1024,6144,0.011003555523024665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1024,7168,0.01752355529202355
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,7168,0.14153244760301378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,7168,0.010464000205198923
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,1536,256,0.005848888721730974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1024,6144,0.016453334026866488
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,6144,0.010025777750545079
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,6144,0.1411448849572076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1024,5120,0.015456888410780164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1024,4096,0.007968000239796108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,5120,0.009313777916961247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,5120,0.14186488257514104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1024,3584,0.0069751110341813825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1024,4096,0.01480977733929952
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,4096,0.008319999608728621
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,4096,0.14635377460055882
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1024,3584,0.014322666658295525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,3584,0.007948444121413762
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,3584,0.1389511161380344
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1024,2560,0.0059004442559348206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1024,3072,0.013767999907334646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,3072,0.14127111434936523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1024,2560,0.013463999662134381
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1024,2048,0.006963555183675554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,2560,0.007314666277832455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,2560,0.1471217738257514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,12288,0.14791644944085017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1024,2048,0.01309866706530253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1024,1536,0.006256888724035687
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,2048,0.1422533326678806
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1024,1536,0.012483555409643384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,1536,0.006698666761318843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1024,1024,0.00453955555955569
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,1536,0.14865333504147
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1024,5120,0.009333333207501305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1024,768,0.00416088890698221
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1024,1024,0.012100444071822695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,1024,0.006286222073766921
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,1024,0.14024799399905735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1024,512,0.0034977777136696708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1024,768,0.012096000214417776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,768,0.005973333285914526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,768,0.14299466874864367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1024,512,0.012064889073371887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1024,3072,0.00673333348499404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,512,0.005940444353553984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,512,0.13976089159647623
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1024,128,0.0031519999934567344
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1024,256,0.01144444445768992
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,256,0.005605333381228977
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,256,0.14145155747731528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,3072,0.007646222081449296
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1024,64,0.002850666642189026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1024,32,0.0031635556370019913
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,128,0.005601777798599667
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,1024,128,0.132076448864407
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1024,64,0.011424889167149862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,768,65536,0.022670222653283015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1024,32,0.012028444144460889
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,768,16384,0.012276444170210095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,1024,2048,0.006973333656787872
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,768,65536,0.08311911424001057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,65536,0.0581279993057251
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,768,12288,0.010694222317801582
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,65536,0.21175200409359404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,16384,0.017017778423097398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,16384,0.14626488420698378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,768,12288,0.02197422252760993
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,12288,0.014097778333557976
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,12288,0.1406631072362264
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,768,10240,0.020087111327383254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,10240,0.14062578148312038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,10240,0.01275111072593265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,1024,256,0.0032257777121331957
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,768,8192,0.01824711097611321
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,8192,0.011086222198274402
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,8192,0.14433066050211588
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,768,7168,0.010084444450007545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,1024,128,0.011677333050303988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,768,7168,0.01718133356836107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,7168,0.010665777656767102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,7168,0.14313599798414442
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,768,6144,0.01644622286160787
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,6144,0.010042666561073726
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,768,5120,0.00868000007337994
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,768,5120,0.015506666567590503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,768,16384,0.028544889556037054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,5120,0.14182310634189182
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,768,4096,0.007918222082985772
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,5120,0.009353777600659264
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,768,10240,0.010216000179449717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,768,4096,0.014781332678265043
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,4096,0.14175289207034641
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,768,3584,0.0069297779765393995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,4096,0.00868711123863856
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,768,3072,0.006545777950021956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,768,3584,0.01442488862408532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,3584,0.00795733349190818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,768,8192,0.011041777829329172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,3584,0.14208000236087376
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,768,2560,0.005597333527273602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,768,3072,0.014111111561457316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,3072,0.007635555333561367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,3072,0.1432008875740899
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,768,2048,0.005244444641802046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,768,2560,0.013728000223636627
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,2560,0.007655111451943715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,2560,0.14477777481079102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,768,1536,0.004558222161398994
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,768,2048,0.01311288856797748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,2048,0.14830666118197972
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,768,6144,0.009500444763236577
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,768,1024,0.003871110992299186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,768,1536,0.012506666282812754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,1536,0.006643555644485686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,1536,0.14388622177971735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,768,768,0.0038204445607132385
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,6144,0.13988888263702393
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,1024,0.006254222244024277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,768,1024,0.012422222230169507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,768,512,0.0035208890007601846
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,768,768,0.012396444049146442
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,768,0.006225777582989798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,768,0.14218578073713514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,768,256,0.003154666680428717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,512,0.005953777581453323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,512,0.1448720031314426
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,768,128,0.0028231110837724474
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,768,256,0.011420444481902652
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,256,0.14070933394961888
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,256,0.005648889061477449
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,768,64,0.0028364445186323593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,768,128,0.011376889215575324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,128,0.005731555736727185
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,128,0.13314399454328749
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,768,32,0.002542222125662698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,512,65536,0.022044445077578228
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,768,64,0.011767110890812345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,768,32,0.011408000356621213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,512,16384,0.012129777835475074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,512,65536,0.08180889156129625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,65536,0.057116442256503634
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,65536,0.20916710959540474
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,512,16384,0.027735999888843958
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,16384,0.016857778032620747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,768,2048,0.006978666616810693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,512,12288,0.02165777815712823
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,12288,0.014135999812020196
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,512,10240,0.009903999666372934
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,12288,0.14577066898345947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,768,1024,0.14915911356608072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,512,8192,0.00904444439543618
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,10240,0.01278133359220293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,512,10240,0.020238222347365487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,10240,0.14071910911136204
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,512,7168,0.008283555507659912
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,8192,0.011221333510345884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,512,8192,0.018067555295096505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,8192,0.14634754922654894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,512,7168,0.017077333397335477
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,512,6144,0.007636444436179266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,7168,0.01037599974208408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,512,5120,0.008389333056079017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,6144,0.010019555687904358
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,6144,0.14032355944315592
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,512,5120,0.015139554937680563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,5120,0.14399200015597874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,512,4096,0.007295110987292395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,5120,0.009279111193286048
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,512,12288,0.01106844428512785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,16384,0.14565511544545492
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,512,4096,0.0148044443792767
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,4096,0.1486319965786404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,512,3584,0.0069679998689227635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,4096,0.008395555946562026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,512,3072,0.006272000157170826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,512,3584,0.014466666513019137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,3584,0.00795288880666097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,3584,0.14757422606150308
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,512,2560,0.005919999960396025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,512,3072,0.013752889302041797
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,3072,0.007722666694058313
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,768,512,0.012039110892348819
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,3072,0.14581067032284206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,512,2048,0.005214222189452913
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,512,2560,0.01313244468635983
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,2560,0.007305777735180325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,2560,0.14247643947601318
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,512,2048,0.013439999686347114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,512,1536,0.004530666602982415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,2048,0.007002666592597961
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,2048,0.1416320006052653
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,512,1024,0.004041777716742622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,7168,0.1441671053568522
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,1536,0.006587555425034628
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,512,1536,0.012448888685968188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,512,6144,0.016248888439602323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,1536,0.1453582180870904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,512,1024,0.012412444584899478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,512,512,0.003177777760558658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,1024,0.006241777704821692
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,1024,0.14834133783976236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,512,768,0.012048888537618848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,768,0.006261333409282897
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,512,256,0.0028817777832349143
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,768,0.14021866851382786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,512,0.005904888941182031
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,512,128,0.002846222163902389
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,512,0.1421768930223253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,512,256,0.011744889120260874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,256,0.005897777775923411
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,256,0.14146488242679173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,512,128,0.01109155515829722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,512,128,0.00564533347884814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,512,128,0.13323999775780573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,512,64,0.011376889215575324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,256,65536,0.018232888645595975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,512,32,0.011868444581826528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,256,65536,0.07806399795744154
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,65536,0.20842844910091826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,65536,0.05079466766781277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,256,16384,0.026358221968015034
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,16384,0.14556888739267984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,256,12288,0.009667555491129557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,16384,0.016865778300497267
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,512,768,0.0038684445122877755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,256,12288,0.02184177769554986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,12288,0.014078223043017916
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,256,10240,0.00905688852071762
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,12288,0.14231467247009277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,512,512,0.012139555480745105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,256,8192,0.008252444366614023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,10240,0.01274222218328052
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,256,10240,0.019911999503771465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,10240,0.14540178245968288
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,512,64,0.0027875554644399216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,256,7168,0.007891555627187094
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,8192,0.011133333047231039
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,256,8192,0.018235555953449674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,8192,0.1387653350830078
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,256,7168,0.01718311177359687
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,7168,0.010656000011497073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,256,6144,0.008630221916569604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,7168,0.1403777731789483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,256,5120,0.007961778177155389
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,256,6144,0.016467556357383728
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,6144,0.010015111002657149
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,6144,0.13946578237745497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,256,4096,0.007292444507280986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,5120,0.1410951084560818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,256,16384,0.0099973330895106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,256,4096,0.01479644411140018
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,4096,0.13825866911146376
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,256,3584,0.014152889450391134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,3584,0.007982222570313348
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,256,3072,0.0064151109092765385
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,3584,0.14762400256262884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,3072,0.007655111451943715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,256,3072,0.014058666096793281
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,3072,0.1418773333231608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,256,2560,0.006280000011126201
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,256,2560,0.013474666410022311
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,2560,0.0074106670088238185
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,2560,0.1417893303765191
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,256,2048,0.005192888693677055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,256,2048,0.013168888787428537
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,512,32,0.0028488888508743713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,256,1536,0.004560889055331548
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,2048,0.0069715554515520734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,256,1536,0.012409778104888068
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,256,5120,0.015576889117558798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,256,1024,0.003866666720973121
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,256,3584,0.006956444846259222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,1536,0.006675555474228329
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,256,1024,0.0123848890264829
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,256,768,0.003495111233658261
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,1024,0.006275555739800136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,1024,0.14059910509321424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,256,768,0.012154666913880242
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,768,0.006210666563775804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,256,512,0.003156444471743372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,768,0.14772266811794704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,256,512,0.012177777787049612
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,512,0.005936000082227919
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,256,256,0.00315022220214208
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,512,0.14250844054751927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,256,256,0.011238222320874533
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,256,128,0.0027706666539112725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,256,0.14340799384646946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,2048,0.14130844010247126
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,256,128,0.011054221954610614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,128,0.1342195537355211
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,256,64,0.002847111059559716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,128,0.005598222215970357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,5120,0.009319999979601966
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,256,64,0.011653333074516721
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,128,65536,0.015237333046065437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,256,1536,0.14425955878363714
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,256,32,0.01142133358452055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,4096,0.008316444853941599
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,128,16384,0.009371555513805812
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,128,65536,0.07738844553629558
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,65536,0.20763199859195283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,16384,0.1422826713985867
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,128,12288,0.00832088871134652
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,16384,0.01681688924630483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,128,12288,0.021589333812395733
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,12288,0.145961774720086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,128,10240,0.008796444369686974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,12288,0.014089778065681458
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,10240,0.012792888614866467
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,128,10240,0.01993688941001892
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,10240,0.14126133918762207
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,128,8192,0.008002666963471307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,256,256,0.005592888842026393
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,128,7168,0.007667555577225155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,8192,0.011076444553004371
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,128,8192,0.018241778016090393
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,8192,0.14556888739267984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,128,7168,0.017032000753614638
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,128,6144,0.007298666569921706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,7168,0.14399911297692194
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,7168,0.010674667027261523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,128,5120,0.007253333098358578
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,256,32,0.0028355556229750314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,128,6144,0.016188444362746347
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,6144,0.009986666341622671
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,128,4096,0.0069102222720781965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,5120,0.009362666971153682
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,128,5120,0.015469332536061605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,5120,0.14245066377851698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,128,3584,0.0062977779242727495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,128,4096,0.014589332871966891
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,4096,0.14547999699910483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,4096,0.008223999705579545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,128,3072,0.005944889038801193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,65536,0.05031377739376492
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,128,3584,0.014448889427714877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,128,16384,0.02612977723280589
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,3584,0.14208977752261692
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,128,2560,0.006262222097979651
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,128,3072,0.013457777599493662
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,3072,0.007732444339328342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,128,2048,0.006276444428496891
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,2560,0.007292444507280986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,2560,0.14117688602871364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,128,1536,0.004193777839342753
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,128,2048,0.013426666458447775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,2048,0.0069617778062820435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,2048,0.1480462286207411
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,128,1024,0.0038808890514903595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,128,1536,0.012549333274364471
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,1536,0.006611555400821898
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,128,1024,0.012103111379676394
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,128,768,0.0035377778112888336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,1024,0.006267555471923616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,128,768,0.011747555600272285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,768,0.144031114048428
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,128,512,0.00332266671790017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,768,0.005924444645643234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,128,512,0.012082666986518435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,512,0.1479039986928304
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,3584,0.00797155582242542
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,512,0.0059004442559348206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,3072,0.14419999387529162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,128,256,0.011700444751315646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,128,2560,0.01349511080318027
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,256,0.1433280044131809
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,128,128,0.0028417776856157514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,256,0.00566133318675889
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,128,128,0.011711999773979187
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,128,0.13634222083621556
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,128,64,0.0025884444928831523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,80,128,128,0.005600000007285013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,128,32,0.0025128889828920364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,64,65536,0.014588443769348992
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,128,64,0.011417778001891242
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,1024,0.14176621701982287
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,128,32,0.011747555600272285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,64,16384,0.008393777741326226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,64,12288,0.007931555310885111
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,64,10240,0.007296000089910295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,64,65536,0.07672977447509766
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,64,16384,0.026047110557556152
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,64,12288,0.021632888250880774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,64,7168,0.006263999889294307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,64,10240,0.01994222237004174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,64,6144,0.006263111200597551
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,64,8192,0.01790577835506863
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,64,7168,0.01697688963678148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,64,5120,0.006225777582989798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,64,4096,0.0059004442559348206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,64,6144,0.016183111402723525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,128,256,0.0028640000770489373
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,64,3584,0.006224888894293044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,64,5120,0.015447111593352424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,64,3072,0.006234666539563074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,64,4096,0.014450665977266101
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,64,2560,0.006048888795905643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,64,3584,0.013818666338920593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,64,2048,0.005943999936183293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,64,3072,0.013805333111021252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,64,1536,0.004512000001139111
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,64,1024,0.0038773334688610504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,64,2560,0.013427555561065674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,64,768,0.0034871109657817413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,64,2048,0.01278222186697854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,64,1536,0.012434666355450949
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,64,512,0.003093333293994268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,64,256,0.0028239999794297745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,1536,0.14524711502922907
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,64,768,0.012039110892348819
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,64,128,0.002817777709828483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,64,512,0.011759111450778114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,64,64,0.0028524444335036804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,64,256,0.011039111349317761
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,64,128,0.011392888923486074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,32,65536,0.01419911119672987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,64,64,0.011406222151385413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,64,32,0.01014311114947001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,64,8192,0.0068897778789202375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,80,128,6144,0.14135111702813044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,32,12288,0.007649777664078607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,32,10240,0.0073520003093613525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,32,16384,0.025848888688617285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,32,8192,0.006953777538405524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,32,12288,0.02163200080394745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,32,10240,0.0199635558658176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,32,7168,0.006601777755551868
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,32,8192,0.017817777064111497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,32,6144,0.006261333409282897
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,32,5120,0.006235555642180973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,32,7168,0.01716977854569753
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,32,4096,0.005894222193294101
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,32,6144,0.015829333000712924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,32,3584,0.006248000181383557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,32,4096,0.014837332897716098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,32,3072,0.005903999838564131
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,64,1024,0.01236533373594284
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,32,2560,0.006280000011126201
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,32,3072,0.013847110999955071
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,64,32,0.0024968888610601425
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,32,2048,0.005910222315125995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,32,2560,0.013402666482660504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,32,16384,0.008626666333940294
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,32,1024,0.0038977778620190094
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,32,1536,0.012714666624863943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,32,768,0.0035422220826148987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,32,65536,0.07584355274836223
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,32,512,0.00313688897424274
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,32,1024,0.012079111403889127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,32,256,0.0028648889727062653
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,32,768,0.012079999678664737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,32,128,0.0028248888750871024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,32,512,0.011686222420798408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,32,64,0.002530666689078013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,32,256,0.011401777466138205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,32,32,0.0027866665687825945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,32,128,0.011737777955002256
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,32,64,0.011348444554540845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,32,32,0.011737777955002256
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,32,3584,0.013802666631009845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,65536,16384,0.17907467153337267
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,65536,16384,0.36394309997558594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,80,32,1536,0.004514666481150521
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,65536,12288,0.27485156059265137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,65536,12288,0.13554755846659342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,32,2048,0.013200889031092325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,65536,10240,0.11556266413794623
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,12288,0.11347733603583442
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,80,32,5120,0.01519733336236742
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,10240,0.08970577849282159
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,65536,8192,0.18484089109632704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,12288,0.24367200003729927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,65536,8192,0.0992506676250034
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,16384,0.3125831021202935
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,16384,0.14786133501264784
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,65536,7168,0.162188450495402
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,10240,0.21065778202480742
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,65536,10240,0.2304044564565023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,65536,7168,0.08471200201246475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,8192,0.07065155771043566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,65536,6144,0.13919911119672987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,65536,6144,0.07507999738057454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,7168,0.0687368909517924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,65536,5120,0.11822577317555745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,6144,0.05530666642718845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,65536,4096,0.09567733605702718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,65536,5120,0.06458755334218343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,7168,0.16487377219729954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,5120,0.04705066813362969
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,65536,4096,0.055164443122016064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,65536,3584,0.08413866493437026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,4096,0.03997244437535604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,5120,0.15007021692064074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,65536,3584,0.04998222324583265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,4096,0.14878222677442762
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,3584,0.03604799840185378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,65536,3072,0.07422844568888347
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,65536,2560,0.06260889106326632
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,3584,0.14869956175486246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,3072,0.031705776850382485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,65536,3072,0.04514844550026787
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,65536,2048,0.05089155501789517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,3072,0.17493955294291177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,2560,0.029134223858515423
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,8192,0.17921333842807344
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,65536,2048,0.03546488947338528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,65536,1536,0.039552887280782066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,2560,0.14531289206610784
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,2048,0.02351911034848955
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,65536,1536,0.030360887447992962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,2048,0.14310755994584826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,65536,1024,0.0299271113342709
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,1536,0.018911111685964797
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,65536,768,0.022661333282788593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,1536,0.14167377683851454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,65536,1024,0.024274667104085285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,1024,0.015146666102939181
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,1024,0.1443528864118788
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,65536,768,0.02060622142420875
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,768,0.1408711142010159
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,768,0.013455111119482251
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,65536,256,0.009036444127559662
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,512,0.012053333222866058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,512,0.14296000533633763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,65536,128,0.0069271110826068455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,65536,256,0.015857777661747403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,65536,512,0.017474666237831116
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,256,0.1402240064409044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,65536,64,0.004827555682924059
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,65536,128,0.01482222146458096
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,128,0.1327751080195109
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,65536,2560,0.040066666073269315
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,65536,64,0.015497777197096082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,65536,32,0.01853511068556044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,16384,65536,0.21033955944908989
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,16384,65536,0.35449687639872235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,65536,6144,0.148561782307095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,16384,16384,0.08259466621610853
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,65536,512,0.01603200038274129
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,16384,16384,0.06146488587061564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,16384,12288,0.0646666685740153
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,16384,0.04865511258443197
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,65536,0.15871644020080566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,16384,12288,0.0481608874268002
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,256,0.01088444474670622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,16384,0.147434671719869
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,16384,10240,0.053680890136294894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,65536,128,0.01071911139620675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,12288,0.036716444624794856
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,16384,8192,0.04586933387650383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,16384,10240,0.04240622123082479
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,12288,0.14074756039513484
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,10240,0.03275288807021247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,16384,7168,0.03891999853981866
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,16384,8192,0.03717333409521315
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,8192,0.024648000796635944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,10240,0.14847911728752985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,16384,6144,0.03392444385422601
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,16384,7168,0.03324177861213684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,8192,0.14496533075968424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,7168,0.023748444186316595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,16384,5120,0.029376887612872656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,7168,0.14035644796159533
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,6144,0.02045866681469811
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,16384,6144,0.030509332815806072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,16384,5120,0.027141332626342773
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,6144,0.1460720035764906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,65536,0.2554391225179036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,5120,0.01794488893614875
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,5120,0.1451102230283949
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,16384,3584,0.021571555071406897
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,16384,4096,0.024225778049892847
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,4096,0.14398222499423557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,4096,0.015542222393883599
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,16384,3072,0.019243554936514962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,16384,3584,0.02233244478702545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,3584,0.14056355423397487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,3584,0.014485332700941296
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,16384,2560,0.016200888488027785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,16384,3072,0.020781333247820537
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,3072,0.013222222526868185
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,65536,32,0.004883555488453971
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,16384,2048,0.013213333156373767
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,16384,2560,0.01924888955222236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,2560,0.012661333713266583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,2560,0.1404906643761529
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,16384,1536,0.011005333728260465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,16384,2048,0.018201778332392376
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,2048,0.010987555815113915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,2048,0.1442355579800076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,16384,1024,0.008543111383914948
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,16384,1536,0.016539555456903245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,1536,0.13854132758246526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,16384,768,0.0069120000633928515
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,16384,1024,0.01444533301724328
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,1024,0.008068444828192392
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,1024,0.13960177368587917
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,16384,512,0.005519999812046687
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,768,0.007449777589903937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,16384,768,0.013737777868906656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,16384,4096,0.025722665919197932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,16384,512,0.012631110846996307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,16384,256,0.0038773334688610504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,512,0.0069226668112807805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,512,0.1408275498284234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,16384,128,0.003559999995761447
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,16384,256,0.012258666257063547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,256,0.14774844381544325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,256,0.006627555522653792
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,16384,64,0.0032844444115956626
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,16384,128,0.012031111452314587
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,128,0.00665511108107037
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,128,0.13213245073954263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,16384,32,0.003568888952334722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,16384,64,0.012096000214417776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,3072,0.14756622579362658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,16384,32,0.011791999969217511
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,12288,65536,0.24838667445712617
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,12288,65536,0.18610933091905382
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,12288,16384,0.06748621993594699
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,16384,1536,0.009368889033794403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,12288,12288,0.05319911241531372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,65536,0.13734132713741726
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,16384,0.041950222518708974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,16384,768,0.1451102230283949
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,16384,0.14166933960384792
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,12288,12288,0.04327022367053562
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,12288,10240,0.04411111275355021
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,12288,0.033118221494886614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,12288,0.1456880039638943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,12288,8192,0.03724622064166599
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,12288,10240,0.03839733203252157
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,10240,0.02722666660944621
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,10240,0.14292800426483154
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,12288,8192,0.03426400158140395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,12288,7168,0.03221866819593642
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,8192,0.022279111875428095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,8192,0.1462657716539171
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,12288,7168,0.030216889248953924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,7168,0.02027466727627648
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,65536,0.22189333703782824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,7168,0.145322667227851
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,12288,6144,0.02788977821667989
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,12288,5120,0.02443733314673106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,6144,0.019533332851197984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,12288,5120,0.025199999411900837
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,12288,4096,0.02093422247303857
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,5120,0.14725066555870905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,5120,0.017310221989949543
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,12288,16384,0.054492443799972534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,12288,3584,0.017882666654056974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,12288,4096,0.02201866606871287
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,4096,0.014734221829308404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,4096,0.14529866642422146
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,12288,3072,0.01588088936275906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,12288,3584,0.020657777786254883
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,3584,0.013348444468445249
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,3584,0.139284438557095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,12288,2560,0.013432000246312885
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,12288,3072,0.01957066688272688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,3072,0.012520888613329994
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,12288,2048,0.011581333147154914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,3072,0.14594222439659968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,12288,2560,0.018746667438083224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,2560,0.13997244834899902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,12288,2048,0.017318222257826064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,2048,0.1452124383714464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,2048,0.009698666632175446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,12288,6144,0.027792887555228338
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,12288,1536,0.016137777103318107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,12288,1024,0.007948444121413762
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,1536,0.00867022242810991
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,1536,0.14316888650258383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,12288,768,0.006567111031876669
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,12288,1024,0.013567110730542077
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,1024,0.007648888561460707
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,1024,0.14224889543321398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,12288,512,0.005032000028424793
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,12288,768,0.012737777498033313
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,768,0.007313778003056844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,768,0.1407288842731052
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,12288,512,0.012390221986505719
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,512,0.14402488867441812
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,512,0.006904888898134232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,12288,256,0.012099555797047086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,256,0.0063368889192740125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,256,0.14062933127085367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,12288,128,0.012064889073371887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,128,0.13732977708180746
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,2560,0.011807111402352652
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,12288,128,0.006589333216349284
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,12288,1536,0.009301332963837517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,12288,32,0.003531555748648114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,12288,64,0.012435555458068848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,12288,32,0.012671111358536614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,10240,65536,0.22095466984642875
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,10240,65536,0.17656800482008192
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,12288,6144,0.14305688275231257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,10240,16384,0.054104887776904635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,12288,256,0.0038319999972979226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,10240,16384,0.05219199922349718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,10240,12288,0.04240355557865567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,65536,0.1306426657570733
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,16384,0.038000000847710505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,16384,0.14681155151791042
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,10240,10240,0.03489333391189575
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,10240,12288,0.04187022315131294
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,12288,0.03128355410363939
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,12288,64,0.003168888803985384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,12288,0.14212444093492296
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,10240,10240,0.03743911120626662
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,10240,8192,0.029927998781204224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,10240,0.025742222865422566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,10240,7168,0.025714667307005987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,10240,0.14335732989841038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,65536,0.21052089002397326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,10240,8192,0.033144887950685285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,8192,0.022834666901164587
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,10240,6144,0.022728888524903193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,8192,0.14485600259568956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,10240,7168,0.02914311157332526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,10240,5120,0.019862223002645705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,7168,0.01963555481698778
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,10240,6144,0.026484444737434387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,6144,0.017869333426157635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,6144,0.1416835520002577
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,10240,4096,0.017148445049921673
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,10240,5120,0.023584889041052923
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,5120,0.13780444198184544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,5120,0.016136889656384785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,10240,3584,0.01443555619981554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,10240,4096,0.021254221598307293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,4096,0.14413155449761286
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,10240,3072,0.01296266665061315
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,10240,3584,0.020230222079488967
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,12288,128,0.003334222154484855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,3584,0.013223110801643796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,10240,2560,0.011571555501884885
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,10240,3072,0.019641778535313077
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,3072,0.012278222375445895
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,3072,0.1387582222620646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,10240,2560,0.01833422150876787
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,10240,2048,0.009972444838947719
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,2560,0.14115911059909397
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,2560,0.011193777951929303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,10240,1536,0.008330666356616551
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,10240,2048,0.01715288890732659
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,2048,0.14246755176120335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,10240,1024,0.00665866666369968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,10240,1536,0.014867555763986377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,1536,0.14088088936275905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,1536,0.00850222259759903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,10240,768,0.00545155546731419
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,10240,1024,0.01349511080318027
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,1024,0.007607111500369177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,1024,0.14383378293779162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,7168,0.14124977588653564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,10240,512,0.004117333226733738
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,10240,768,0.01277511070171992
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,768,0.007258666886223688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,768,0.14093244075775146
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,10240,256,0.003547555456558863
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,512,0.006607999818192587
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,512,0.14344799518585205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,10240,256,0.012390221986505719
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,4096,0.01406311160988278
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,10240,128,0.0032035555276605818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,256,0.13950933350457087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,256,0.006612444503439798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,10240,64,0.003198222153716617
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,3584,0.1444524394141303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,128,0.006579555571079254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,10240,128,0.1338133282131619
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,10240,32,0.0032097777972618737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,10240,64,0.012080888781282636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,10240,32,0.01274311128589842
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,8192,65536,0.1802355580859714
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,10240,2048,0.009336888790130615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,8192,65536,0.12395289209153916
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,8192,16384,0.05036177900102404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,8192,16384,0.04063733418782552
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,65536,0.07891733116573758
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,16384,0.02535111043188307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,8192,12288,0.033471110794279314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,16384,0.1460968918270535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,8192,12288,0.033152888218561806
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,12288,0.019575999842749704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,10240,512,0.012443555725945367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,12288,0.1422951089011298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,8192,10240,0.029600001043743555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,8192,8192,0.023732443650563557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,65536,0.2050622171825833
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,10240,0.017087999317381117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,8192,8192,0.025763556361198425
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,8192,7168,0.02129688858985901
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,10240,128,0.011755555868148804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,8192,0.14148088296254477
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,8192,0.014890667464998035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,8192,6144,0.01872711049185859
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,8192,7168,0.022679110368092854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,7168,0.014085332552591959
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,7168,0.14308799637688532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,8192,6144,0.02128177715672387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,6144,0.012739555703269111
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,6144,0.13950844605763754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,8192,4096,0.01441422270403968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,8192,5120,0.019865777757432725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,5120,0.011796444654464722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,5120,0.14317688677046034
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,8192,4096,0.018208000395033095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,8192,3584,0.013091555900043912
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,4096,0.14271556006537542
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,4096,0.010681777364677854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,8192,3072,0.011726222104496427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,8192,3584,0.017496888836224873
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,3584,0.1441671053568522
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,3584,0.009712888962692684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,8192,2560,0.01036444471942054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,8192,3072,0.01682933337158627
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,3072,0.009008000294367472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,8192,2048,0.009337777892748514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,8192,2560,0.01612711118327247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,2560,0.009003555609120263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,2560,0.14157422383626303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,10240,0.14918666415744358
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,8192,2048,0.01497511068979899
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,2048,0.008009778128729926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,2048,0.14293866687350804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,8192,1536,0.013460444079505073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,1536,0.14499199390411377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,1536,0.007306666837798224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,8192,5120,0.01626666718059116
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,8192,768,0.004504000147183736
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,1024,0.0069573331210348345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,1024,0.1423679987589518
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,8192,768,0.012075555821259817
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,768,0.0064186664919058485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,768,0.13979110452863905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,8192,512,0.012083555261294046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,512,0.1393244398964776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,8192,256,0.003479111111826367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,512,0.005943999936183293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,3072,0.14099466800689697
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,8192,256,0.011734222372372946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,8192,128,0.00349866681628757
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,256,0.1434266699684991
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,256,0.005936000082227919
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,8192,1536,0.007650666766696506
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,8192,64,0.003154666680428717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,8192,128,0.011743110915025076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,8192,128,0.136007997724745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,8192,32,0.0034968890249729156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,8192,1024,0.00554044461912579
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,8192,64,0.011704888608720569
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,8192,32,0.0118568887313207
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,7168,65536,0.15652799606323242
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,7168,65536,0.12025600009494358
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,7168,16384,0.04593422346644931
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,8192,512,0.0038355555799272326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,65536,0.07535466882917616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,7168,16384,0.038310223155551486
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,7168,12288,0.03567733367284139
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,16384,0.023795555035273235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,16384,0.14119556215074328
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,8192,10240,0.02846933404604594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,7168,12288,0.031162665949927434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,7168,10240,0.026901332868470088
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,12288,0.14542578326331243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,12288,0.018921777606010437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,7168,10240,0.02760266595416599
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,65536,0.20416888925764295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,7168,8192,0.023447111248970032
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,10240,0.016453334026866488
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,10240,0.14579644468095568
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,7168,8192,0.0237128883600235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,8192,0.01458577811717987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,8192,0.14240354961819118
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,7168,7168,0.021721776988771226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,7168,6144,0.017493334081437852
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,7168,0.01350400017367469
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,7168,0.14313333564334446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,8192,1024,0.01240799989965227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,7168,6144,0.020390222469965618
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,6144,0.012530666258600024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,6144,0.1405751175350613
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,7168,4096,0.013172444370057849
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,7168,5120,0.018952000472280715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,5120,0.011468444433477191
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,5120,0.14557154973347983
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,7168,4096,0.017878222796652053
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,7168,3584,0.011630222201347351
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,4096,0.14804711606767443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,7168,3584,0.01716088917520311
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,7168,3072,0.011267555256684622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,3584,0.009191110730171204
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,3584,0.1438995599746704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,7168,3072,0.016512889001104567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,7168,2560,0.009986666341622671
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,3072,0.14503822061750624
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,7168,7168,0.019917334119478863
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,7168,2560,0.015623999966515435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,8192,128,0.005952888892756567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,2560,0.14654489358266196
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,7168,2048,0.008636444807052612
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,2560,0.00832799987660514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,7168,1536,0.007262222468852997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,7168,2048,0.014523555835088095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,2048,0.14662310812208387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,7168,5120,0.015111999379263984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,7168,1536,0.013094222380055321
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,7168,1024,0.005192888693677055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,1536,0.14114133516947427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,1536,0.0069511110583941145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,7168,768,0.004556444370084339
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,7168,1024,0.012479110724396177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,1024,0.14213777912987605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,1024,0.006360888895061281
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,4096,0.009696000152164036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,7168,768,0.012133333418104382
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,768,0.006239111224810283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,768,0.1450808842976888
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,7168,256,0.0035084444615576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,7168,512,0.01204533295498954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,512,0.005952888892756567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,512,0.14579199420081243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,7168,128,0.0031893334041039148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,7168,256,0.012078222301271228
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,3072,0.008635555704434713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,256,0.13949422041575113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,256,0.0058746664888328975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,7168,64,0.003137777869900068
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,7168,128,0.011392888923486074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,128,0.005895111295912001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,7168,128,0.13688533835940891
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,7168,64,0.01202222208182017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,7168,32,0.010710222025712332
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,6144,65536,0.12541955047183567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,7168,2048,0.007302222152551015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,6144,65536,0.11056177483664618
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,6144,16384,0.03909066650602553
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,6144,16384,0.03612355391184489
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,65536,0.07287910911771986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,6144,12288,0.03267466690805223
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,16384,0.021756443712446425
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,16384,0.14519643783569336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,6144,12288,0.028899557060665552
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,6144,10240,0.02793955471780565
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,12288,0.017847999930381775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,12288,0.1420782274670071
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,6144,8192,0.020431111256281536
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,65536,0.20302489068773058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,10240,0.01609688831700219
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,6144,10240,0.025207999679777358
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,10240,0.14362488852606878
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,6144,8192,0.02273599969016181
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,6144,7168,0.020336889558368258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,6144,6144,0.017829333742459614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,8192,0.13937332895067003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,6144,7168,0.02085777786042955
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,7168,0.013103110922707452
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,7168,0.14715822537740073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,6144,5120,0.016456888781653512
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,7168,32,0.0031582222630580268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,6144,6144,0.01959111127588484
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,6144,0.1409413284725613
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,6144,4096,0.014160000615649752
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,6144,5120,0.018211555149820115
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,5120,0.14235022332933214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,5120,0.010882666541470421
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,6144,3584,0.012705778082211813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,6144,4096,0.017483555608325534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,4096,0.009480888644854227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,6144,3584,0.01647111111217075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,3584,0.008623110751310984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,6144,3072,0.009645333720578088
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,6144,3072,0.015834665960735746
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,7168,512,0.003887999802827835
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,3072,0.1504782173368666
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,6144,2560,0.008607111043400234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,3072,0.008290666672918532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,6144,2560,0.014767110347747803
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,6144,2048,0.007631999750932057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,2560,0.14138844278123644
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,2560,0.007983999947706858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,8192,0.013859555953078799
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,6144,1536,0.006230222268237009
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,6144,2048,0.013758222262064615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,2048,0.007269333634111617
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,2048,0.13948444525400797
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,6144,1024,0.004554666578769684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,6144,0.012116444607575735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,1536,0.006931555353932911
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,6144,1536,0.013010666602187686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,1536,0.1443155606587728
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,6144,768,0.004182222402758068
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,6144,1024,0.012438221938080259
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,1024,0.006311999840868845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,1024,0.14574310514662
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,6144,512,0.0037342223028341928
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,6144,768,0.012158222496509552
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,4096,0.14583110809326172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,768,0.14267022079891628
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,6144,256,0.0031848889258172777
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,6144,512,0.011708444191349877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,512,0.005902222047249476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,3584,0.13811555173661974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,512,0.1416986650890774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,6144,128,0.002926222152180142
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,6144,256,0.01205244412024816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,6144,64,0.0028826666788922418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,256,0.1456773281097412
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,6144,32,0.002830222249031067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,128,0.005903111149867375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,6144,128,0.011519110865063138
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,6144,128,0.13265511724683973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,6144,64,0.011737777955002256
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,6144,32,0.011716444459226398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,5120,65536,0.11763289239671494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,5120,16384,0.0337777766916487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,5120,65536,0.10384888781441583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,65536,0.07086222039328681
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,5120,16384,0.033534222178988986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,5120,12288,0.028762665059831407
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,16384,0.021423111359278362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,5120,12288,0.02686400049262577
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,12288,0.017432888348897297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,5120,10240,0.024345777100986902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,65536,0.20076889461941191
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,12288,0.14371466636657715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,5120,8192,0.02072622213098738
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,768,0.006248000181383557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,5120,10240,0.024711110525661047
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,10240,0.14976978302001953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,5120,7168,0.01923733287387424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,8192,0.013668444421556262
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,8192,0.1405457788043552
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,6144,256,0.005632888939645555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,5120,6144,0.014558222558763293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,7168,0.012734221915404001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,7168,0.1411315600077311
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,5120,6144,0.018896889355447557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,6144,0.011743110915025076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,5120,5120,0.01313066648112403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,5120,5120,0.017895110779338412
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,5120,0.1465839942296346
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,16384,0.1451902257071601
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,5120,4096,0.01109333336353302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,5120,0.010371555884679159
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,5120,3584,0.010100444157918295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,5120,4096,0.016928889685206942
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,4096,0.14854578177134195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,5120,3584,0.016167110866970487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,3584,0.008328888979223039
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,5120,3072,0.009328888522254096
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,10240,0.015616888801256815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,5120,8192,0.021828444467650518
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,5120,3072,0.01514755520555708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,5120,2560,0.007972444097201029
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,3072,0.14225599500868055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,3072,0.00795733349190818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,5120,2048,0.007250666618347168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,5120,7168,0.020004444652133517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,5120,2560,0.013787555197874704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,2560,0.007606222397751278
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,2560,0.1432808902528551
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,5120,1536,0.005895111295912001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,5120,2048,0.013092444174819522
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,2048,0.007260444263617198
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,6144,0.1414533323711819
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,5120,1536,0.012805333567990197
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,5120,1024,0.004523555437723796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,1536,0.141649776034885
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,1536,0.006918222126033571
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,5120,768,0.00416088890698221
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,5120,1024,0.012432888978057437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,1024,0.006339555399285422
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,1024,0.13972889052497015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,5120,512,0.0038364442686239877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,5120,768,0.012087110843923358
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,768,0.006228444476922353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,768,0.14262310663859049
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,4096,0.00923911150958803
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,5120,512,0.012048000262843238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,512,0.0059004442559348206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,512,0.13898666699727377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,5120,128,0.003148444410827425
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,5120,256,0.012039999994966718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,3584,0.1452488899230957
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,256,0.1420560015572442
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,256,0.005894222193294101
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,5120,64,0.003156444471743372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,5120,32,0.003142222141226133
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,5120,128,0.011709333293967776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,5120,128,0.005944889038801193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,128,0.13113955656687418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,5120,64,0.011413333316644033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,4096,65536,0.08743821912341648
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,5120,32,0.011413333316644033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,4096,16384,0.02851288848453098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,4096,65536,0.09647999869452582
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,4096,16384,0.03166044420666165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,65536,0.06868177652359009
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,5120,2048,0.14529511663648817
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,4096,12288,0.023024888502226934
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,16384,0.020607110526826646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,16384,0.13959556155734593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,4096,10240,0.018738667170206707
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,4096,12288,0.02444000045458476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,12288,0.016855110724767048
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,12288,0.1481013298034668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,65536,0.20045688417222765
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,4096,8192,0.01628266606065962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,4096,10240,0.022583999567561682
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,10240,0.14417066839006212
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,4096,7168,0.014087999860445658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,8192,0.013070222404268054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,8192,0.13987199465433756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,5120,256,0.0032115555885765287
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,4096,6144,0.0125937776433097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,7168,0.14051644007364908
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,4096,6144,0.018160889546076458
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,6144,0.011164444188276926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,4096,5120,0.01147377739350001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,4096,5120,0.017095110482639737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,5120,0.1434017817179362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,4096,4096,0.010436444646782346
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,5120,0.010002666877375709
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,4096,4096,0.016168889072206285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,4096,3584,0.009054222040706212
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,4096,0.1469013293584188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,4096,0.008664000365469191
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,4096,3584,0.01520088811715444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,3584,0.007993777592976889
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,4096,3072,0.008303110798199972
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,10240,0.015106666419241162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,4096,3072,0.014122666584120857
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,3072,0.14336621761322021
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,4096,2560,0.00756533361143536
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,3072,0.00795377790927887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,4096,7168,0.019395555059115093
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,4096,2048,0.006361777583758037
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,4096,2560,0.014091556270917257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,2560,0.14025599426693386
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,2560,0.007609777980380588
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,4096,1536,0.005218666460778978
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,6144,0.13873777124616835
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,4096,2048,0.013103110922707452
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,2048,0.007304000357786815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,2048,0.14275199837154812
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,4096,1024,0.004184000194072723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,4096,1536,0.012729778057999082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,1536,0.006829333388143116
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,1536,0.13911378383636475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,4096,768,0.0038568890757030914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,4096,1024,0.01238933371173011
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,1024,0.1401831176545885
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,4096,512,0.003495111233658261
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,4096,768,0.01237422227859497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,768,0.14504267109764948
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,4096,512,0.011758222348160215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,4096,256,0.003192000091075897
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,512,0.13936532868279353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,512,0.005948444621430502
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,3584,0.14818755785624185
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,4096,256,0.011734222372372946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,4096,128,0.0028604444944196274
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,256,0.005631999837027655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,4096,8192,0.02088977727625105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,4096,128,0.011369778050316704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,128,0.13921688662634954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,7168,0.012190221912331052
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,4096,32,0.0028035555862718155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,4096,64,0.011055111057228513
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,4096,32,0.011399110986126794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3584,65536,0.07649422354168363
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,1024,0.006284444282452266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,768,0.005942222144868638
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3584,16384,0.025387555360794067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,65536,0.06755999724070232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,65536,0.20098400115966797
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,16384,0.14256889290279812
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,4096,256,0.1396177742216322
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3584,12288,0.020344000723626878
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,16384,0.020271999968422782
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3584,12288,0.02496266696188185
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,4096,64,0.0028568889117903183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,4096,128,0.005897777775923411
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,12288,0.01681066718366411
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,12288,0.14271377192603216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3584,8192,0.01540177729394701
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3584,10240,0.022318222456508215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,10240,0.01512888901763492
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3584,8192,0.020612445142534044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3584,7168,0.013600889179441663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,8192,0.14089955223931208
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3584,65536,0.09423200289408366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3584,7168,0.01885688967174954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3584,6144,0.012107555237081317
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,7168,0.14074844784206814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,7168,0.011686222420798408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3584,6144,0.018019555343521964
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,6144,0.010670222342014313
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,6144,0.1401217778523763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3584,16384,0.0303457776705424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3584,5120,0.017092444830470614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,5120,0.13954844739702013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3584,4096,0.015492444237073263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3584,10240,0.01789066692193349
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,4096,0.13966400093502468
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,4096,0.00830488900343577
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,10240,0.1443964507844713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3584,3584,0.014456000592973499
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,8192,0.012729778057999082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,3584,0.14397688706715903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3584,3072,0.00795555528667238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,3584,0.008265777594513362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3584,3072,0.014156444205178155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,3072,0.14773421817355686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3584,2560,0.007261333366235097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,3072,0.007700444095664554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3584,5120,0.010713777608341642
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3584,2560,0.013420444395807056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3584,2048,0.005932444499598608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,2560,0.007435555259386699
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,2560,0.144031114048428
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3584,1536,0.004933333231343163
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3584,4096,0.010020444790522257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,2048,0.0069751110341813825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,2048,0.14313066005706787
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3584,1024,0.004179555508825514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3584,1536,0.01273333364062839
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,1536,0.006738666858938005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,1536,0.14493689272138807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3584,768,0.003666666646798452
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3584,1024,0.012349333200189801
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,1024,0.13852177725897893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,1024,0.00629066675901413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3584,512,0.003544888976547453
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3584,768,0.012424888710180918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,768,0.006268444574541516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,768,0.1405493285920885
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3584,256,0.0031724443866146933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3584,3584,0.008541333178679148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3584,512,0.011745777395036487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,512,0.005961777849329843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3584,256,0.011807999677128263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,256,0.005883555445406172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,256,0.1410035557217068
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3584,64,0.0028444443725877335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3584,128,0.011753777662913004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3584,32,0.002834666727317704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,128,0.005927111125654644
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,128,0.1320613357755873
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3584,64,0.011391110718250275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3584,32,0.011709333293967776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3072,65536,0.09104710817337036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3072,16384,0.02162222233083513
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,65536,0.06637689140107897
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,3584,5120,0.00959022260374493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3072,16384,0.029096000724368628
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,65536,0.20029955440097383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,16384,0.01995377739270528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,16384,0.14587644735972086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3072,12288,0.023690667417314317
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,12288,0.14270844724443224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3072,10240,0.015765332513385348
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3584,128,0.0029893333299292457
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,12288,0.01648622254530589
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3072,10240,0.02180622186925676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,10240,0.014561777313550314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,10240,0.14416089322831896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3072,8192,0.02017955482006073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3072,65536,0.06808444526460436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,8192,0.14456000592973497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,8192,0.0123848890264829
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3072,7168,0.0185244447655148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,7168,0.13967377609676784
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3584,2048,0.012801777985360889
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3072,6144,0.013081777426931592
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,7168,0.010948444406191507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3072,12288,0.018165333403481376
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3072,5120,0.010402667025725046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,6144,0.010020444790522257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,6144,0.1418915589650472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3072,5120,0.016522667474216886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,5120,0.009699555734793345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3072,4096,0.009394666386975182
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,3584,512,0.14011643992529974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3072,4096,0.015099555253982544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,4096,0.1422355572382609
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,4096,0.008531555533409119
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3072,8192,0.014069333672523499
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3072,3584,0.014496000276671516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3072,3072,0.007644444704055786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,3584,0.13979378011491564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3072,7168,0.013097777962684631
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3072,3072,0.014078223043017916
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,3072,0.1409120029873318
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3072,2560,0.006792888873153263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,3072,0.00793511089351442
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3072,2560,0.01351288871632682
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,2560,0.007592889169851939
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,2560,0.1416755517323812
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3072,2048,0.0058782220714622075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3072,6144,0.017529777354664274
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3072,2048,0.012768000364303589
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3072,1536,0.004912000149488449
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,2048,0.14106488227844238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3072,1024,0.004253333227501975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3072,1536,0.012444444828563266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,1536,0.006996444529957241
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,5120,0.15024267302619085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3072,768,0.0038951109680864546
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3072,1024,0.012359999948077731
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,1024,0.006251555350091722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,1024,0.1475084490246243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3072,3584,0.008260444634490544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3072,768,0.01204533295498954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3072,512,0.003530666646030214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,768,0.14080444971720377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,768,0.00628266649113761
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3072,256,0.0032799999333090256
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3072,512,0.011769777370823754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,512,0.005929777605666055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,512,0.14615821838378906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3072,128,0.002846222163902389
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,3584,0.007998222278224098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3072,256,0.011782222323947482
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,256,0.140937778684828
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3072,64,0.0028515555378463534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,256,0.00556355549229516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,3072,32,0.0030133333057165146
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3072,128,0.011438222395049201
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,128,0.00591111100382275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,128,0.13904889424641928
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2560,65536,0.05901955233679878
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3072,64,0.011426666544543372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,3072,32,0.011711999773979187
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2560,16384,0.020071110791630216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2560,65536,0.08724355697631836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2560,16384,0.028930667373869155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,65536,0.06494400236341688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,16384,0.01977155605951945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2560,12288,0.01647733317481147
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2560,10240,0.014879110786649915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,3072,2048,0.007136888802051544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2560,12288,0.02335111134582096
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,65536,0.2004071076711019
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,12288,0.016167999969588388
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,12288,0.1476666662428114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2560,8192,0.013225778109497495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2560,10240,0.02125688890616099
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,10240,0.14080799950493708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2560,7168,0.01239377756913503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2560,8192,0.01962311069170634
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,8192,0.011720889144473605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,8192,0.1477866702609592
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2560,6144,0.01147555559873581
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2560,7168,0.018615111708641052
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,7168,0.01109688894616233
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,7168,0.14634044965108237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2560,6144,0.017228444417317707
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2560,5120,0.010713777608341642
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,6144,0.1412595510482788
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,6144,0.010103111465771994
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2560,4096,0.009002666506502364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2560,5120,0.015503111812803479
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,5120,0.1461297803454929
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,5120,0.009304000271691216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2560,3584,0.008082666330867344
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2560,4096,0.014599111345079211
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,4096,0.14534578058454725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2560,3584,0.014451555079884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,3584,0.14036088519626194
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2560,3072,0.007304000357786815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,3584,0.007986666427718269
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,16384,0.14283466339111328
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2560,3072,0.01384800010257297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,3072,0.007648888561460707
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2560,2560,0.006303999986913469
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,3072,0.14268622133466932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2560,2048,0.00553688903649648
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,2560,0.007325333025720384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2560,2560,0.013731555806265937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,2560,0.1446275578604804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,3072,1536,0.1399822235107422
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2560,2048,0.013111111190583972
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,2048,0.007152000235186682
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,2048,0.14212444093492296
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2560,1024,0.0041742221348815495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2560,1536,0.012712000144852532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,1536,0.14172977871365017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,1536,0.00657066661450598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2560,768,0.0037626665499475268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2560,1024,0.012097777591811286
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,1024,0.1398906707763672
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2560,768,0.012375999655988483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2560,512,0.0035297779573334586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,768,0.14133689138624403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,4096,0.008366222182909647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2560,512,0.012071111136012606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,512,0.14046043819851345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,512,0.005932444499598608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2560,256,0.011747555600272285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,256,0.14729155434502497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,256,0.005610666755172942
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2560,128,0.00314666661951277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,10240,0.014123555686738757
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2560,1536,0.004855111241340637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2560,128,0.01164533363448249
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2560,64,0.002809777855873108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,2560,128,0.1389333274629381
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2560,32,0.0028657778683635923
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,128,0.005889777921968036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2560,64,0.011375111010339526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2560,32,0.010701333483060202
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2048,16384,0.01889955500761668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,1024,0.00628266649113761
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2048,65536,0.08635466628604466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,65536,0.06324444214502971
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2048,16384,0.027034666803148057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2048,12288,0.015019555886586508
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,16384,0.14070399602254233
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,65536,0.19913155502743193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2048,10240,0.013623110949993134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2048,12288,0.02315644423166911
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,12288,0.0151555554734336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,12288,0.1492017772462633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2048,10240,0.02091111077202691
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,10240,0.013058666553762225
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,10240,0.14494666788313124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2048,8192,0.018976888722843595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2048,7168,0.011321778098742167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,8192,0.011389333340856763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2048,7168,0.017508443858888414
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,7168,0.14590222305721706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2048,6144,0.010405333505736457
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2048,65536,0.047888888253106006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,7168,0.010681777364677854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2048,6144,0.016865778300497267
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,6144,0.010049777726332346
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,6144,0.14390399720933703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,2560,768,0.006219555520349079
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2048,5120,0.015459555718633862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2048,4096,0.008681777450773451
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,5120,0.14397688706715903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,5120,0.009288888838556077
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2560,256,0.0033084443873829315
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2048,4096,0.01443733274936676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,4096,0.14479555024041071
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2048,8192,0.012764444781674279
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2048,3584,0.01386044422785441
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,3584,0.007968000239796108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,3584,0.1444853279325697
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,8192,0.14239911238352457
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2048,3072,0.013744889034165276
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,3072,0.14326133992936876
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2048,2560,0.006255110932721033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,3072,0.007657777931955125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2048,2560,0.013418667018413544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,2560,0.1479840013715956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2048,2048,0.005604444278611078
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2048,5120,0.010085333552625444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2048,2048,0.01274400038851632
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,2048,0.14431999789343938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2048,1536,0.004552888787455029
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,16384,0.019304000669055514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,2048,0.006955555743641323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2048,1536,0.012676444318559436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,1536,0.006600889066855113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,1536,0.13946844471825495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2048,3584,0.0076453329788313965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,4096,0.00833422193924586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2048,768,0.0038133333954546186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2048,1024,0.012084444363911947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,1024,0.00628266649113761
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,1024,0.14112799697452122
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2048,512,0.003547555456558863
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2048,3072,0.0069013333155049225
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2048,768,0.01203555530971951
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,768,0.006088888893524806
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,768,0.14654755592346191
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2048,256,0.0031537777847713898
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2048,512,0.012028444144460889
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,512,0.005895111295912001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,512,0.13886222574445936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2048,128,0.00286311118139161
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2048,256,0.011563555234008364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,256,0.13899732960595026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2048,128,0.011640888949235281
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,2560,0.007298666569921706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,2048,128,0.1330311165915595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2048,32,0.002867555452717675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,128,0.005888888819350137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1536,65536,0.038578665918774076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2048,64,0.011358222199810876
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,2048,32,0.011410666836632622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1536,16384,0.015089778436554803
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1536,65536,0.0826551119486491
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,65536,0.06254044506284925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1536,16384,0.02646933330429925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,65536,0.20032444265153673
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1536,12288,0.013168888787428537
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,16384,0.018713777263959248
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,12288,0.014144000079896716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,12288,0.1410399940278795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1536,10240,0.011524444652928246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1536,10240,0.020896888441509668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,10240,0.14064978228674993
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1536,8192,0.01071466671095954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,10240,0.013043555948469373
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,2048,256,0.005612444546487596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1536,8192,0.018294221825069852
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1536,7168,0.010335110955768162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,8192,0.14111822181277806
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,8192,0.011069333387745751
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1536,7168,0.016999999682108562
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,7168,0.01038577738735411
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1536,6144,0.009648000200589498
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,7168,0.1405493285920885
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1536,5120,0.008949333594905006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,6144,0.010031999813185798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1536,6144,0.016484444340070088
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,6144,0.14493155479431152
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1536,5120,0.015106666419241162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1536,4096,0.008488888541857401
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,5120,0.009310222334331935
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,5120,0.1412444379594591
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1536,3584,0.00794488853878445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1536,12288,0.022653333014912073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,4096,0.008344000412358178
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1536,3584,0.014084445105658637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1536,3072,0.006680888848172293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,3584,0.1421004401312934
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1536,3072,0.013762666947311826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,3072,0.14093155331081814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1536,2560,0.005955555372767978
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,3072,0.007659555309348636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1536,2560,0.013526221944226159
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,2560,0.1425857808854845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1536,2048,0.005212444398138259
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,2560,0.007298666569921706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1536,2048,0.012849777936935425
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,16384,0.14932355615827772
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,2048,0.14328622817993164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1536,1536,0.004580444345871608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,2048,0.0069875551594628235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1536,4096,0.0144177774588267
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1536,1536,0.01277955538696713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,4096,0.14527555306752524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,1536,0.006631111105283101
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1536,1024,0.0123857781291008
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,1024,0.006253333141406377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1536,768,0.003539555602603488
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1536,768,0.012059555285506778
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,768,0.14123822583092585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1536,512,0.0031964443624019623
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,768,0.006115555349323485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1536,512,0.012081777883900536
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,512,0.14132711622450086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,512,0.00591466658645206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1536,256,0.012068444656001197
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2048,1024,0.0038515557017591265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,256,0.14091911580827501
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1536,1024,0.004008000095685323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,1536,0.14688621626959905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1536,128,0.00281866660548581
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,256,0.005582222094138463
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,3584,0.00794044468137953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,1024,0.1435182227028741
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1536,64,0.0028124445428450904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1536,32,0.0025271111064487034
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,1536,128,0.00554933316177792
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1024,65536,0.0314062237739563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1536,64,0.011445333560307821
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1536,32,0.011361777782440186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,2048,64,0.0027866665687825945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1024,16384,0.012422222230169507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1024,65536,0.07966578006744385
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,65536,0.058486223220825195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1024,16384,0.02602400051222907
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1024,12288,0.010423111418883005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1536,256,0.0031537777847713898
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,16384,0.01681511104106903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1024,12288,0.021873777111371357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,12288,0.014121777481502958
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1024,10240,0.010652444428867765
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,12288,0.1424657768673367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1024,10240,0.01939022209909227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,10240,0.14096800486246744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1024,8192,0.009296000003814697
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1024,8192,0.01753422286775377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,8192,0.14000177383422852
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1536,128,0.011065777805116443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,1536,128,0.13217066393958196
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,8192,0.011186666786670685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1024,7168,0.01649777756796943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,7168,0.1382711066140069
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1024,6144,0.008296888735559251
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,7168,0.010661332971519895
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1024,6144,0.015848888291252982
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,6144,0.0099973330895106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,6144,0.14749066034952799
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1024,5120,0.00831911133395301
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,65536,0.19922044542100695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1024,4096,0.007343110938866933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,16384,0.14279822508494058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1024,5120,0.01538222200340695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,5120,0.15024177233378092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1024,3584,0.0069439998931354946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1024,4096,0.014462222655614218
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,4096,0.008279999925030602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,4096,0.14328000280592176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1024,3072,0.006568888823191325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1024,3584,0.013791110780504016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,3584,0.007993777592976889
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,3584,0.14454221725463867
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,10240,0.01276266657643848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1024,2560,0.005929777605666055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1024,3072,0.01387733303838306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,3072,0.007599111232492659
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,3072,0.14652800559997559
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1024,2048,0.005241777747869492
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1024,2560,0.013117333253224691
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,2560,0.1439359982808431
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,2560,0.007292444507280986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1024,1536,0.004514666481150521
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1024,2048,0.012728888955381183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,2048,0.006976000136799282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,2048,0.14059200551774767
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1024,1024,0.0038977778620190094
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1024,1536,0.01276000009642707
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,1536,0.006595555692911148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,1536,0.14433511098225912
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1024,768,0.0035333335399627686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1024,1024,0.01236533373594284
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,1024,0.006228444476922353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,1024,0.14115911059909397
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1024,512,0.0034755555291970572
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1024,768,0.012117333710193634
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,768,0.006228444476922353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,768,0.1477244430118137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1024,256,0.0028799999919202593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1024,512,0.011749332977665795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,512,0.14203022585974798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,512,0.005919999960396025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,5120,0.009349333743254343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1024,128,0.0028577778074476454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1024,256,0.011393778026103973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,256,0.005672889037264719
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1024,64,0.002855111120475663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1024,32,0.002537777854336633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1024,128,0.011737777955002256
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,1024,128,0.005604444278611078
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,128,0.13139377699957952
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,768,65536,0.025582222474945918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1024,64,0.011367999845080905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,1024,32,0.01036711119943195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,768,65536,0.0773786637518141
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,65536,0.058318224218156606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,768,12288,0.00980177770058314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,65536,0.19870666662851968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,16384,0.14744444688161215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,16384,0.01685333251953125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,1024,7168,0.008799110849698385
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,768,12288,0.020996444755130347
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,768,10240,0.01034755590889189
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,12288,0.014124444789356656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,768,10240,0.01958044370015462
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,10240,0.01277955538696713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,10240,0.14191999700334337
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,768,8192,0.008994667066468133
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,8192,0.011383111278216044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,8192,0.14123733838399252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,768,7168,0.008228444390826756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,768,7168,0.01683644453684489
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,1024,256,0.13967733912997776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,7168,0.14511111047532824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,768,6144,0.008271111382378472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,7168,0.010431999961535135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,768,6144,0.015857777661747403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,768,16384,0.011323555476135679
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,768,5120,0.008316444853941599
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,6144,0.14225867059495714
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,6144,0.009988444546858469
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,768,4096,0.007599111232492659
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,768,16384,0.02478755513827006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,768,5120,0.015138667490747241
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,5120,0.009342222577995723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,5120,0.14479822582668728
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,768,3584,0.007280888656775157
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,768,4096,0.014383110735151501
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,4096,0.008291555775536431
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,4096,0.14465778403811985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,768,3072,0.0069217777086628815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,768,3584,0.01425066590309143
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,3584,0.00797333319981893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,3584,0.13965688811408147
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,768,2560,0.006893333461549547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,768,3072,0.013585777746306526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,3072,0.14135644170973036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,768,2560,0.013437333206335703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,768,2048,0.005212444398138259
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,2560,0.14221955670250788
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,2560,0.007314666277832455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,768,8192,0.017553778158293832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,768,1536,0.00453244439429707
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,768,2048,0.013056000073750814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,2048,0.14080088668399388
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,768,1024,0.0038942222793896994
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,768,1536,0.012725333372751871
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,1536,0.00664533343580034
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,1536,0.14317156208886042
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,768,768,0.0035075553589397003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,768,1024,0.012104888757069906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,1024,0.14359822538163927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,768,768,0.01203644441233741
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,768,512,0.003170666595300039
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,768,0.005943999936183293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,768,0.1423457728491889
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,768,512,0.011765333513418833
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,512,0.14313333564334446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,12288,0.13983999358283147
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,768,256,0.011784888803958893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,3072,0.007679111427730984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,256,0.13885955015818277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,768,128,0.0028391112056043413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,256,0.005602666487296422
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,768,128,0.011394666300879585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,128,0.005577777822812398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,768,128,0.13738932874467638
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,2048,0.006909333169460297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,768,32,0.002521777732504739
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,512,65536,0.019596444235907663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,768,64,0.011751111182901593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,768,32,0.011050666371981302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,1024,0.006230222268237009
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,512,65536,0.07427822219000922
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,512,16384,0.01106844428512785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,65536,0.05748089154561361
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,65536,0.20002488295237222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,512,12288,0.008991999758614434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,512,16384,0.024679111109839544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,16384,0.14105955759684244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,512,10240,0.008965333302815756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,512,12288,0.021024000313546922
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,768,512,0.0058906666106647914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,12288,0.014131555954615274
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,12288,0.1471102237701416
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,768,256,0.002936888900068071
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,512,8192,0.00832711077398724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,512,10240,0.01942133406798045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,10240,0.0127368892232577
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,10240,0.14369955327775744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,512,7168,0.007617777420414819
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,512,8192,0.01719288859102461
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,8192,0.011151110960377587
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,8192,0.14480533864763048
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,512,6144,0.007281777759393056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,512,7168,0.016588444511095684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,7168,0.010663111176755695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,7168,0.1435848871866862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,512,6144,0.01587022178702884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,6144,0.009970666633711921
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,6144,0.1434897714191013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,512,4096,0.007255111303594377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,512,5120,0.01516533394654592
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,768,64,0.0028035555862718155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,5120,0.14324711428748235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,512,3584,0.0069439998931354946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,512,4096,0.014139556222491793
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,4096,0.008270222279760573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,4096,0.1398879951900906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,512,3072,0.006623999940024481
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,512,3584,0.014116444521480136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,3584,0.00793511089351442
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,3584,0.14346667130788168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,512,2560,0.006600889066855113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,512,3072,0.013768889009952545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,3072,0.14124266306559244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,3072,0.007653333246707916
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,512,2048,0.006269333263238271
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,16384,0.01684888866212633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,2560,0.14068978362613255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,2560,0.007380444142553542
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,512,1536,0.004517333375083076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,512,2048,0.01278222186697854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,2048,0.006953777538405524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,2048,0.137990223036872
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,512,1024,0.003846222327815162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,512,1536,0.012723555167516073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,1536,0.006731555693679386
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,1536,0.1404213375515408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,512,768,0.0035377778112888336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,512,1024,0.012427555190192329
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,512,5120,0.007962666451931
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,1024,0.1424773269229465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,512,512,0.003145777723855443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,512,768,0.012079999678664737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,768,0.006227555374304454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,768,0.14136621687147352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,5120,0.009038222332795462
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,512,256,0.0028693332440323303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,512,512,0.011905777785513135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,512,0.005906666732496685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,512,0.14069067107306585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,512,128,0.002543111021320025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,512,256,0.011156444748242697
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,256,0.005610666755172942
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,256,0.13975021574232313
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,512,64,0.002508444504605399
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,512,32,0.002518222149875429
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,512,128,0.011367110742463006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,128,0.005591111050711737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,512,128,0.13277244567871094
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,256,65536,0.015443555182880826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,512,64,0.011372444530328115
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,256,16384,0.009970666633711921
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,512,2560,0.013232000172138214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,65536,0.1986924409866333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,256,16384,0.024333332975705464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,256,12288,0.008504889077610439
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,16384,0.016875555117925007
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,256,12288,0.02095199955834283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,12288,0.14155733585357666
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,256,10240,0.007682667010360294
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,12288,0.01407199932469262
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,256,10240,0.01923111081123352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,10240,0.1423004468282064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,256,8192,0.007250666618347168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,10240,0.012771555946932899
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,256,8192,0.017492444978819955
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,8192,0.14299999343024358
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,256,7168,0.006924444602595435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,8192,0.011090666883521609
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,512,32,0.011396444506115384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,256,65536,0.07213955455356173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,256,7168,0.016661332713233102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,256,6144,0.006653333289755716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,7168,0.14077511098649767
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,7168,0.010602666272057427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,256,5120,0.006226666685607698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,256,6144,0.015792888071801927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,16384,0.14907733599344888
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,6144,0.009977777798970541
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,6144,0.1466577715343899
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,256,5120,0.015284443895022074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,5120,0.009364444348547194
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,5120,0.14270844724443224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,256,3584,0.0063884444534778595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,512,1024,0.006233777850866318
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,4096,0.1430471075905694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,256,3584,0.013941332697868347
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,3584,0.007983110845088959
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,3584,0.14031110869513616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,256,3072,0.013763555222087435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,3072,0.007657777931955125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,256,2560,0.006574222197135289
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,3072,0.1424471139907837
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,256,2560,0.013429332938459186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,2560,0.007325333025720384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,2560,0.14291022883521184
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,256,2048,0.006213333457708359
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,65536,0.05065244436264038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,256,2048,0.013064888616402945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,256,1536,0.004497777670621872
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,2048,0.007239110767841339
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,256,1536,0.0127688886390792
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,1536,0.13976177904340956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,256,1024,0.0038408889538711975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,256,4096,0.014125333891974555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,4096,0.008279999925030602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,256,1024,0.012359999948077731
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,256,768,0.003452444449067116
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,256,3072,0.006278222219811545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,1024,0.14643911520640054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,1024,0.0062871111763848206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,256,512,0.003165333428316646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,256,768,0.012072000238630505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,768,0.005940444353553984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,768,0.14362488852606878
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,256,256,0.003176888864901331
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,256,512,0.011763555308183035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,512,0.005601777798599667
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,512,0.13926933871375188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,256,128,0.002500444443689452
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,256,256,0.011739555332395764
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,256,0.14464710818396673
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,256,128,0.011406222151385413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,128,0.005568000177542369
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,256,64,0.0025155554629034465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,128,0.13585155540042454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,256,2048,0.1407395601272583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,256,32,0.0027351110345787476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,128,65536,0.014785778191354541
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,256,64,0.011372444530328115
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,256,32,0.011410666836632622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,256,4096,0.006621333460013072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,128,16384,0.008014221986134848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,128,65536,0.07118844323688082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,65536,0.1994311147265964
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,1536,0.006616888774765863
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,128,12288,0.007592000067234039
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,128,16384,0.024093333217832778
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,16384,0.016798221402698092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,16384,0.14535644319322374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,128,10240,0.007228444847795699
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,128,12288,0.020961778031455144
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,12288,0.014096000128322177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,12288,0.1475484503640069
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,128,8192,0.006959111326270633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,10240,0.14023999373118082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,10240,0.012732444538010491
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,128,7168,0.006632888896597757
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,128,8192,0.017288888494173687
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,8192,0.011222222612963783
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,8192,0.14019733005099827
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,128,6144,0.0063324446479479475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,128,7168,0.016872889465755887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,7168,0.010491555763615502
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,7168,0.14230310916900635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,128,5120,0.006286222073766921
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,128,6144,0.015899555550681222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,6144,0.009696889254781935
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,6144,0.14275199837154812
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,128,4096,0.006232000059551663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,5120,0.14175732930501303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,5120,0.009347555538018545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,128,4096,0.014502222339312235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,4096,0.00832088871134652
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,128,3584,0.006298666612969504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,4096,0.14058132966359457
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,65536,0.05022844341066149
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,128,3584,0.013790222505728403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,3584,0.14230667220221624
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,128,3072,0.013739555246300168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,128,10240,0.019440889358520508
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,3072,0.1420160002178616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,128,2560,0.006272888845867581
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,3072,0.007660444411966536
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,128,2560,0.013461333182122974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,128,2048,0.005936000082227919
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,256,256,0.005605333381228977
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,128,2048,0.013100444442696042
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,128,1536,0.0041964443193541635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,2048,0.006934222247865465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,128,5120,0.01510933372709486
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,128,1536,0.01276533305644989
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,128,1024,0.003823111040724648
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,1536,0.006761777732107375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,1536,0.14648000399271646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,128,1024,0.012434666355450949
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,128,768,0.0035004446076022256
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,1024,0.006028444402747684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,1024,0.14050755235883924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,128,3072,0.005955555372767978
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,128,512,0.0032035555276605818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,768,0.005944889038801193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,128,768,0.012091555529170565
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,768,0.14767821629842123
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,128,256,0.002883555574549569
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,128,512,0.011726222104496427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,512,0.00591555568906996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,512,0.14078756173451742
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,128,256,0.011058666639857821
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,128,128,0.0028497777465316984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,256,0.005609777652555042
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,2560,0.007301333049933116
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,128,0.005603555589914322
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,128,128,0.01072177787621816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,128,64,0.002516444358560774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,128,32,0.0028133332315418455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,128,64,0.01072177787621816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,2048,0.14056089189317492
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,128,32,0.01107911103301578
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,64,65536,0.010822222464614444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,64,16384,0.007307555940416124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,64,12288,0.006629333313968446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,64,65536,0.06969955232408312
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,64,16384,0.023690667417314317
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,64,10240,0.006535111202134027
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,64,12288,0.020896888441509668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,64,128,3584,0.00796622203456031
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,64,10240,0.018830221560266282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,64,8192,0.017216889394654166
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,64,6144,0.006268444574541516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,64,7168,0.016162667009565566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,64,5120,0.005938666562239329
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,64,6144,0.015809777710172866
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,64,4096,0.005720888988839255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,256,0.14399377504984537
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,2560,0.14350577195485434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,64,5120,0.014755555325084262
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,64,3584,0.006239999913507038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,64,3072,0.005631999837027655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,64,2560,0.005950222412745158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,64,4096,0.014135110709402295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,64,2048,0.005972444597217772
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,64,3072,0.01348088930050532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,64,1536,0.004229333251714706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,64,2560,0.012860444684823355
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,64,1024,0.003579555700222651
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,64,2048,0.012771555946932899
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,64,768,0.0035013332962989807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,64,1536,0.01239377756913503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,64,512,0.003162666741344664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,64,1024,0.011715555356608497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,64,256,0.002558222247494592
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,64,768,0.012080888781282636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,64,128,0.002847111059559716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,64,512,0.011755555868148804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,64,8192,0.006609777609507243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,64,64,0.0024764444679021835
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,64,256,0.011056889262464313
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,64,32,0.0028080000645584534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,64,128,0.011697777443461947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,32,65536,0.010728889041476779
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,32,16384,0.007133333219422235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,64,32,0.010017777482668558
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,32,12288,0.006607999818192587
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,32,65536,0.06792977783415052
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,32,16384,0.024047111471494038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,32,10240,0.006291555447710886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,32,8192,0.006618666566080517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,32,10240,0.018933332628673978
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,32,8192,0.01718933383623759
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,64,3584,0.014479999740918478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,32,6144,0.006613333192136552
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,32,6144,0.01587733295228746
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,32,5120,0.006587555425034628
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,32,5120,0.015154666370815702
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,32,4096,0.005921777751710679
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,64,7168,0.006636444479227066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,32,3584,0.00611200018061532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,32,4096,0.014465777410401238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,32,3072,0.006259555617968242
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,32,3584,0.014095111025704278
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,32,2560,0.006223111102978389
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,32,12288,0.020625778370433383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,32,3072,0.013752889302041797
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,32,2048,0.005620444400442972
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,32,1536,0.004510222209824456
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,32,1024,0.003543111185232798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,32,2560,0.013412444127930535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,32,7168,0.006264888991912206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,32,2048,0.01310222182008955
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,32,1536,0.012080888781282636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,32,512,0.003145777723855443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,32,256,0.0028373334142896864
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,32,1024,0.012392000191741519
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,32,7168,0.01661688917213016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,32,128,0.002542222125662698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,32,512,0.011366222467687396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,32,256,0.011761777930789523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,32,32,0.002504000026318762
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,32,128,0.011517333487669626
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,32,64,0.010660444696744284
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,32,32,0.011384000380833944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,65536,16384,0.1770951085620456
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,65536,16384,0.36517333984375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,64,64,0.011355555719799466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,65536,12288,0.27336621284484863
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,65536,12288,0.13389689392513698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,32,768,0.0035093331502543557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,16384,0.14952266216278076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,64,32,768,0.01200355589389801
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,64,32,64,0.0025235555238193935
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,65536,10240,0.23037510448031953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,65536,10240,0.11417688263787164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,12288,0.11799911657969157
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,65536,8192,0.18580622143215605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,65536,8192,0.09796088933944702
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,10240,0.08655822277069092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,16384,0.30530667304992676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,65536,7168,0.16266666518317327
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,12288,0.23994578255547416
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,10240,0.20968622631496855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,8192,0.17931911680433485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,65536,6144,0.13878756099277073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,7168,0.07369510995017158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,65536,6144,0.0744088888168335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,6144,0.054720001088248364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,65536,5120,0.117484450340271
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,65536,5120,0.06461777951982287
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,8192,0.0702800022231208
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,6144,0.14842932754092747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,65536,4096,0.09652177492777507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,5120,0.04677866564856636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,65536,7168,0.08392622073491414
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,65536,4096,0.05487111210823059
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,65536,3584,0.08444622490141128
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,65536,3584,0.04991111159324646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,4096,0.1419795619116889
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,7168,0.16253511110941568
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,65536,3072,0.07265066438251071
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,3584,0.04166577921973334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,3584,0.14567111598120794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,65536,2560,0.06167199876573351
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,3072,0.03051911128891839
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,65536,3072,0.045069333579805165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,65536,2560,0.04068177938461304
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,65536,2048,0.049843556351131864
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,2560,0.026851556367344324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,65536,2048,0.035742223262786865
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,2560,0.14307110839419895
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,2048,0.022861333356963262
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,65536,1536,0.038804445001814104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,2048,0.14180089367760554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,5120,0.14779555797576904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,65536,1536,0.030541333887312148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,65536,1024,0.029389331738154095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,1536,0.018883556127548218
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,4096,0.03900266687075297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,65536,768,0.021908443835046556
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,1536,0.14928355481889513
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,65536,1024,0.023981332778930664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,1024,0.014936000108718872
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,1024,0.14385067092047796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,65536,768,0.020253333780500624
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,768,0.013716444373130798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,768,0.14551022317674425
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,65536,256,0.009496000077989366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,65536,512,0.017645332548353408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,512,0.01202044470442666
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,512,0.14344088236490884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,65536,128,0.007638221813572778
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,65536,256,0.015823110938072205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,256,0.010664888554149203
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,256,0.14708711041344538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,65536,128,0.014818666709793938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,128,0.1332346730762058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,65536,32,0.006729777902364731
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,65536,128,0.010777778095669217
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,65536,64,0.0158897770775689
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,65536,32,0.01925244430700938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,16384,65536,0.2031315565109253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,16384,65536,0.34549421734280056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,16384,16384,0.08292888932757907
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,16384,16384,0.059417777591281466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,65536,0.16017422411176893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,65536,512,0.015192889504962497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,16384,12288,0.06363466713163587
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,16384,12288,0.04697333441840278
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,16384,0.04288622074657016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,16384,10240,0.05296088920699226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,65536,64,0.006775110960006714
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,12288,0.03433155682351854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,16384,10240,0.041356444358825684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,12288,0.14060178067949083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,65536,3072,0.1419315603044298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,16384,8192,0.04481777879926893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,10240,0.03380444314744737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,16384,8192,0.03719288773006863
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,10240,0.14079377386305067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,8192,0.02471911079353756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,16384,7168,0.038689778910742864
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,8192,0.14040177398257786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,16384,7168,0.03279022375742594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,7168,0.02236977716286977
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,65536,0.24413866466946074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,7168,0.14061866866217718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,16384,6144,0.03021866745418972
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,6144,0.022257778379652236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,16384,5120,0.02927466564708286
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,6144,0.14054755369822183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,16384,5120,0.02751644452412923
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,16384,4096,0.02499288817246755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,5120,0.1403262217839559
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,5120,0.01787911189926995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,16384,3584,0.021606221795082092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,16384,4096,0.024710221423043147
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,16384,0.1453804439968533
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,4096,0.14159377415974936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,16384,3072,0.01867911054028405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,4096,0.015180443723996481
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,3584,0.0159582214223014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,16384,3072,0.020657777786254883
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,3072,0.14072267214457193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,3072,0.013244444297419654
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,16384,2560,0.01603911154799991
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,16384,2560,0.01994044416480594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,16384,6144,0.03382488754060533
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,2560,0.01310222182008955
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,2560,0.14514756202697754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,16384,2048,0.013436444103717804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,16384,1536,0.011027555498811932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,2048,0.011004444625642566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,16384,2048,0.018485332528750103
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,16384,1536,0.017024889588356018
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,16384,1024,0.0083137775460879
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,1536,0.009321778184837764
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,1536,0.14592444896697998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,16384,1024,0.014545778433481852
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,16384,768,0.007104000283612146
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,1024,0.15032354990641275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,1024,0.008031110796663497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,16384,768,0.013365333278973898
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,16384,3584,0.02199466692076789
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,768,0.14913333786858454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,3584,0.14103378189934626
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,768,0.007623111208279927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,16384,512,0.005423111220200856
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,16384,256,0.0037902221083641052
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,16384,512,0.012421333127551608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,512,0.14828266037835017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,512,0.0069715554515520734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,16384,128,0.003156444471743372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,16384,256,0.012390221986505719
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,256,0.006634666687912411
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,256,0.1450079944398668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,16384,64,0.0031884445084465873
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,16384,128,0.012404444317022959
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,16384,32,0.0031742221779293488
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,128,0.13613422711690268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,16384,128,0.0069448889957533936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,16384,64,0.012647999657524956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,16384,32,0.013108444710572561
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,16384,2048,0.1398222181532118
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,12288,16384,0.0743466681904263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,12288,65536,0.25992798805236816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,12288,65536,0.17326666249169242
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,12288,16384,0.052412443690829806
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,12288,12288,0.053673777315351695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,65536,0.14103466934627956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,12288,12288,0.04218488931655884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,16384,0.04232266545295715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,12288,10240,0.043856001562542386
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,16384,0.141903109020657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,12288,0.1450764470630222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,12288,0.03339733348952399
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,12288,10240,0.037500444385740496
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,12288,8192,0.036995556619432234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,10240,0.029436445898479883
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,12288,8192,0.03320711188846164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,12288,7168,0.0319164428446028
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,10240,0.1407182216644287
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,8192,0.14862577120463052
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,8192,0.024442666106753882
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,12288,6144,0.028061333629820082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,12288,7168,0.030061334371566772
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,7168,0.02241244415442149
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,7168,0.14097689257727727
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,12288,5120,0.02425333360830943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,12288,6144,0.027618666489919026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,65536,0.2137155532836914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,6144,0.018357333209779527
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,12288,4096,0.020900444851981267
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,6144,0.14863644705878362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,5120,0.01706044375896454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,12288,5120,0.0240320000383589
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,12288,3584,0.017864000466134813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,12288,4096,0.02161866592036353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,12288,3072,0.015824000040690105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,4096,0.14074844784206814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,12288,3584,0.0203475554784139
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,3584,0.014187556174066333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,3584,0.14053422874874538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,12288,2560,0.014016888207859464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,12288,3072,0.01957066688272688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,3072,0.012431999875439538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,3072,0.14002133740319148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,12288,2560,0.018560889694425795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,2560,0.011387555963463254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,2560,0.14319199985928005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,12288,2048,0.01754666699303521
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,2048,0.1403297715716892
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,12288,1536,0.009903999666372934
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,2048,0.010031999813185798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,12288,1536,0.01637066735161675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,1536,0.14084800084431967
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,12288,1024,0.00795288880666097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,12288,1024,0.013968888256284924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,1024,0.007607111500369177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,5120,0.13920977380540636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,12288,768,0.007248889240953658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,4096,0.014440889159838358
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,768,0.14170222812228733
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,12288,768,0.01310400002532535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,768,0.007289778027269576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,12288,2048,0.012030222349696688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,12288,512,0.012426666915416718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,512,0.14694400628407797
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,512,0.006700444552633498
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,12288,256,0.011741333537631564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,256,0.14501689540015325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,1536,0.008658666577604082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,256,0.006540444576077991
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,12288,128,0.012103111379676394
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,1024,0.1412471135457357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,12288,64,0.0031999999450312722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,12288,128,0.006592888798978593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,12288,32,0.0032088889016045462
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,12288,64,0.012149333126015134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,12288,32,0.012892444100644855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,12288,512,0.0052284445199701525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,10240,65536,0.21866133477952746
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,12288,256,0.004017777740955353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,10240,65536,0.16992089483473036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,10240,16384,0.06251466936535306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,10240,16384,0.05147555470466614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,12288,128,0.0035297779573334586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,65536,0.13870488272772893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,16384,0.03979111048910353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,16384,0.1460053390926785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,10240,12288,0.041123555766211614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,12288,128,0.13861689302656385
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,10240,10240,0.03524355424775018
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,12288,0.1412471135457357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,10240,10240,0.03674044542842441
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,10240,8192,0.030272887812720403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,10240,0.025368889172871906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,10240,0.14549599753485784
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,10240,7168,0.025669333007600572
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,65536,0.20158844523959688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,8192,0.02161066730817159
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,8192,0.14205422666337755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,10240,6144,0.022679110368092854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,10240,7168,0.028772443532943726
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,7168,0.019713777634832595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,7168,0.14124266306559244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,10240,5120,0.019520888725916546
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,10240,6144,0.025440888272391424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,6144,0.01774133410718706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,6144,0.14644800292121038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,10240,4096,0.017247110605239868
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,10240,5120,0.023286221755875483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,5120,0.016154666741689045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,5120,0.14179466830359563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,10240,12288,0.0421635541650984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,10240,3584,0.014455111490355598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,4096,0.014116444521480136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,4096,0.1400924391216702
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,10240,3072,0.013017777767446307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,10240,3584,0.02017955482006073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,12288,0.03027555677625868
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,3584,0.14119466145833334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,3584,0.014111111561457316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,10240,2560,0.012867555850081973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,10240,3072,0.019241778386963736
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,3072,0.012067555553383298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,3072,0.1411448849572076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,10240,2048,0.011432888607184092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,10240,2560,0.018742221924993727
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,2560,0.14030933380126953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,2560,0.010616000327799054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,10240,1536,0.00998044427898195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,2048,0.009355555805895064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,10240,2048,0.017249777913093567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,10240,1024,0.007802666889296637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,10240,1536,0.015081778168678284
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,1536,0.14038311110602483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,1536,0.008472888833946651
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,10240,768,0.005934222290913264
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,10240,1024,0.013764444324705334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,1024,0.14126843876308864
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,1024,0.00758844448460473
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,64,128,128,0.13748266961839464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,10240,768,0.013099555340078143
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,10240,512,0.005126222140259213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,768,0.006956444846259222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,768,0.14326043923695883
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,10240,256,0.0041662222809261745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,512,0.006605333338181178
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,10240,4096,0.02123111155298021
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,512,0.14654578102959528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,10240,256,0.012089778151777057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,10240,128,0.0035422220826148987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,256,0.1400622261895074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,256,0.006249777972698212
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,10240,128,0.012079111403889127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,128,0.1376711130142212
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,10240,32,0.0035253332720862497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,10240,64,0.012439111040698158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,10240,32,0.013099555340078143
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,10240,8192,0.03271289004219903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,8192,65536,0.17732889122433135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,10240,2048,0.13935022883945042
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,8192,16384,0.04933511217435201
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,8192,65536,0.11916089057922363
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,8192,16384,0.03914755582809448
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,65536,0.07808355490366618
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,8192,12288,0.03751555416319106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,16384,0.024523556232452393
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,10240,512,0.012413333687517377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,8192,12288,0.0318177772892846
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,12288,0.019075555933846366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,10240,64,0.0031795555518733132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,12288,0.14036444822947183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,10240,128,0.006621333460013072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,65536,0.19847910934024385
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,8192,8192,0.02625777820746104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,8192,10240,0.02882933285501268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,10240,0.017151999804708693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,10240,0.14608444107903376
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,8192,7168,0.02279288901223077
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,8192,0.014891554911931356
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,8192,0.14127111434936523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,8192,6144,0.02020088831583659
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,8192,7168,0.022642667094866436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,7168,0.013928888572586907
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,7168,0.13930043909284803
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,8192,5120,0.017362667454613578
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,8192,6144,0.02113688819938236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,6144,0.012737777498033313
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,6144,0.14791999922858343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,8192,5120,0.01941777765750885
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,8192,4096,0.014838222000333997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,5120,0.011767110890812345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,5120,0.14115911059909397
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,16384,0.14719200134277344
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,8192,3584,0.013092444174819522
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,8192,4096,0.01822933389080895
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,4096,0.1463973389731513
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,8192,3072,0.011482666763994427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,4096,0.010438222024175854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,8192,3584,0.017525333497259352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,3584,0.010019555687904358
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,3584,0.1410213311513265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,8192,10240,0.03091288937462701
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,8192,2560,0.010392888552612728
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,8192,3072,0.016778666112158034
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,3072,0.14188533359103733
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,8192,2048,0.00888355573018392
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,8192,2560,0.016019556257459853
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,2560,0.008448888858159384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,2560,0.14330489105648464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,8192,1536,0.007867555651399825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,8192,2048,0.015455111861228943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,2048,0.00793511089351442
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,2048,0.1463279989030626
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,8192,8192,0.025044444534513686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,8192,1024,0.006152888966931238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,8192,1536,0.014099554883109199
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,1536,0.0069653333889113525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,1536,0.14100800620185003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,8192,768,0.005088888936572605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,8192,1024,0.012791111237472959
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,1024,0.006638222270541721
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,8192,512,0.004006222304370669
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,1024,0.1495937771267361
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,8192,768,0.0123848890264829
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,768,0.14125600126054552
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,8192,256,0.0035208890007601846
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,8192,512,0.012082666986518435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,512,0.1471724510192871
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,512,0.005958222266700532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,8192,128,0.0031902222997612427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,8192,256,0.011679111255539788
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,256,0.0058959999846087555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,256,0.14227733347151014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,8192,64,0.0029688889367712867
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,8192,32,0.0028684443483750024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,8192,128,0.011427555647161273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,128,0.005931555396980709
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,8192,128,0.13889155122968885
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,8192,64,0.01181422256761127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,8192,32,0.012409778104888068
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,7168,16384,0.04682755470275879
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,7168,65536,0.11236000061035156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,7168,16384,0.03720444440841675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,3072,0.008985777695973715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,65536,0.07521955834494697
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,7168,12288,0.03726044297218323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,16384,0.02313511073589325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,16384,0.13999288611941868
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,7168,12288,0.03010666701528761
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,7168,10240,0.028435554769304063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,12288,0.018553778529167175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,12288,0.1414008935292562
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,7168,10240,0.02679288884003957
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,7168,8192,0.024664888779322307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,10240,0.016511110795868766
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,10240,0.14077866077423096
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,8192,768,0.006272888845867581
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,65536,0.19721689489152697
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,7168,7168,0.02139555580086178
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,7168,8192,0.023567999402681988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,8192,0.14209066496955022
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,7168,6144,0.018515555395020377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,8192,0.014426666829321118
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,7168,7168,0.021708443760871887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,7168,0.013440888788965015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,7168,0.1476986673143175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,7168,5120,0.016166221764352586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,6144,0.012602667013804117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,7168,6144,0.020288000504175823
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,7168,4096,0.013605333036846586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,7168,5120,0.018852444158660043
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,5120,0.147690667046441
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,5120,0.011418666276666852
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,7168,4096,0.017862222260899015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,7168,3584,0.01204622205760744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,7168,65536,0.15862311257256403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,4096,0.14061333073510066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,7168,3072,0.011272889044549731
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,7168,3584,0.01717333330048455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,3584,0.009180444810125563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,3584,0.14318133725060359
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,7168,2560,0.009876444107956356
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,3072,0.008575999902354347
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,3072,0.14567022853427464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,7168,2560,0.015346666177113851
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,7168,2048,0.008340444829728868
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,2560,0.007641777396202087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,7168,2048,0.014826666977670459
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,2048,0.007303111255168915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,7168,1536,0.007484444313579136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,2048,0.145416882303026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,7168,1536,0.013472889032628802
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,1536,0.0069164443347189165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,1536,0.1422951089011298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,7168,1024,0.00555288874440723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,7168,1024,0.012440000143316058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,1024,0.00628977765639623
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,1024,0.14050132698482937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,7168,768,0.005402666827042897
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,6144,0.1451688872443305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,7168,512,0.0038888889054457345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,7168,768,0.012713777522246042
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,768,0.006227555374304454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,768,0.14115466011895075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,4096,0.010009778042634329
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,7168,256,0.003506666670242945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,7168,512,0.012079111403889127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,512,0.0058871110280354815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,512,0.14191910955641004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,7168,128,0.003154666680428717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,7168,256,0.011742221812407175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,7168,3072,0.016742222838931613
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,256,0.13978311750623915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,256,0.005917333480384614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,7168,32,0.0028524444335036804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,7168,128,0.011695110963450538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,128,0.1357004510031806
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,7168,64,0.011786667009194693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,7168,2560,0.14963911639319524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,7168,32,0.012427555190192329
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,6144,65536,0.13515644603305393
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,6144,16384,0.0387662218676673
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,6144,65536,0.1034897764523824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,65536,0.07284888956281874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,6144,16384,0.03492177857293023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,16384,0.022319111559126113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,6144,12288,0.03176800078815884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,16384,0.14582133293151855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,6144,12288,0.028117332193586562
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,12288,0.018018666240904067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,12288,0.14626222186618382
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,6144,10240,0.024645333488782246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,6144,8192,0.022602667411168415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,10240,0.14719555113050672
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,6144,8192,0.021943110558721755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,7168,64,0.0028657778683635923
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,6144,7168,0.020266667008399963
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,8192,0.014121777481502958
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,7168,128,0.005888000130653381
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,7168,0.01310755560795466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,6144,7168,0.020472000042597454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,7168,0.1441742181777954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,6144,6144,0.015228443675571017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,6144,5120,0.013728000223636627
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,6144,6144,0.019484443797005545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,6144,0.14216444227430555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,6144,5120,0.01794488893614875
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,6144,10240,0.027433777848879497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,5120,0.1409297784169515
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,6144,4096,0.0123831108212471
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,5120,0.011035555766688453
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,65536,0.1960444450378418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,4096,0.14615733093685573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,10240,0.015844444433848064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,6144,3584,0.010668444136778513
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,3584,0.00868444475862715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,3584,0.14368711577521429
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,6144,3072,0.009635555247465769
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,8192,0.1466346714231703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,6144,3072,0.016141333513789706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,3072,0.008283555507659912
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,3072,0.1424880027770996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,6144,2560,0.008380444513426887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,6144,2560,0.014116444521480136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,2560,0.14307199584113225
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,2560,0.007606222397751278
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,6144,0.012178666889667511
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,6144,2048,0.014138667119873894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,6144,1536,0.006583999842405319
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,2048,0.14327200253804526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,2048,0.007241778075695038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,6144,1536,0.013086222112178802
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,1536,0.0069360000391801195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,1536,0.14099289311303034
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,6144,1024,0.0047893333766195504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,6144,4096,0.017440888616773818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,6144,768,0.004184000194072723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,4096,0.009453333086437648
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,1024,0.006454222318198945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,6144,1024,0.012717333104875354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,6144,512,0.003846222327815162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,6144,768,0.01238933371173011
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,768,0.006230222268237009
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,6144,3584,0.01683111157682207
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,6144,256,0.0031840000301599503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,6144,512,0.012048888537618848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,512,0.005895111295912001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,512,0.138991117477417
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,6144,256,0.011755555868148804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,256,0.14255466726091173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,256,0.005888000130653381
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,6144,128,0.011748444702890186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,6144,128,0.005944889038801193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,6144,64,0.002850666642189026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,128,0.1355537838406033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,6144,2048,0.007463110817803278
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,6144,32,0.0028444443725877335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,6144,64,0.011696889168686338
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,6144,32,0.012082666986518435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,5120,65536,0.09614311324225532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,5120,16384,0.03522577881813049
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,65536,0.07043555710050795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,5120,16384,0.032459557056427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,1024,0.14733866850535074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,5120,12288,0.027264000640975103
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,16384,0.14343289534250894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,16384,0.021328889661365088
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,6144,768,0.14147199524773493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,65536,0.19497956169976127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,5120,10240,0.022985777921146814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,5120,12288,0.02540622154871623
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,12288,0.017362667454613578
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,12288,0.14368800322214761
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,5120,8192,0.01994044416480594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,10240,0.015459555718633862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,5120,8192,0.02159999973244137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,5120,7168,0.017992888887723286
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,8192,0.013516444298956128
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,5120,65536,0.11613778273264568
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,5120,7168,0.019711999429596793
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,5120,6144,0.01650311052799225
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,7168,0.14586755964491102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,7168,0.012686221963829465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,5120,6144,0.018900444110234577
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,6144,0.011723555624485016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,5120,5120,0.014089778065681458
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,5120,5120,0.017416000366210938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,5120,0.14487822850545248
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,5120,0.010544000400437249
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,5120,4096,0.01697688963678148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,6144,128,0.003152888889114062
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,4096,0.1392151117324829
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,5120,3584,0.011057777537239922
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,5120,10240,0.02314844396379259
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,4096,0.00905688852071762
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,5120,3584,0.01577422188387977
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,3584,0.008174222376611497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,3584,0.1401475535498725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,8192,0.14498844411638048
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,5120,2560,0.008638222184446123
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,5120,3072,0.015437333120240105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,3072,0.00794044468137953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,3072,0.14324177636040583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,5120,2048,0.007869333028793335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,5120,2560,0.014567999376191033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,2560,0.13929243882497153
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,2560,0.007611555357774098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,5120,2048,0.013452444639470843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,5120,1536,0.005882666756709416
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,2048,0.00702044450574451
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,6144,0.14483288923899332
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,2048,0.1442506710688273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,5120,1024,0.004263111286693149
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,5120,1536,0.012786666552225748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,1536,0.0069164443347189165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,1536,0.13973600334591335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,5120,768,0.003903999924659729
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,5120,4096,0.012401777837011548
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,5120,1024,0.012657778130637275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,1024,0.006239999913507038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,1024,0.14262844456566706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,5120,512,0.003714666598372989
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,768,0.006280888699822956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,768,0.14635022481282553
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,5120,256,0.0032026666320032547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,5120,512,0.011757333245542316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,512,0.005977777971161737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,512,0.14454132980770534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,5120,3072,0.009664889011118147
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,5120,128,0.0032008888406885993
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,5120,256,0.011372444530328115
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,256,0.005790222022268508
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,256,0.14176621701982287
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,5120,64,0.002858666703104973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,5120,32,0.002847111059559716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,5120,128,0.011420444481902652
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,128,0.1352248854107327
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,5120,64,0.011728888584507836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,5120,32,0.012430222498046027
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,4096,65536,0.08885422017839219
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,4096,16384,0.028748446040683325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,65536,0.06818933619393243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,4096,16384,0.029771556456883747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,65536,0.1945271094640096
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,16384,0.14565777778625488
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,16384,0.020394666327370536
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,4096,12288,0.022686221533351477
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,5120,768,0.01238222254647149
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,4096,10240,0.020163555939992268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,4096,12288,0.0236453331179089
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,12288,0.01681600014368693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,5120,10240,0.13966933886210123
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,12288,0.14513244893815783
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,4096,8192,0.017489777670966256
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,10240,0.01517688896920946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,10240,0.1456248892678155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,4096,8192,0.020587555236286588
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,4096,7168,0.015555555621782938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,5120,128,0.005927111125654644
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,8192,0.13936532868279353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,4096,6144,0.013445333474212222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,4096,65536,0.09169066614574856
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,7168,0.012343111137549082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,7168,0.14601511425442165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,4096,5120,0.011186666786670685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,4096,6144,0.018216000662909616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,6144,0.14275111092461482
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,4096,5120,0.01740622189309862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,4096,4096,0.00998577806684706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,5120,0.009640889035330879
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,4096,4096,0.016181333197487723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,4096,3584,0.008941333327028487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,4096,0.008325333396593729
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,4096,0.14594222439659968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,4096,3584,0.01479111115137736
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,3584,0.00812977800766627
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,3584,0.1400924391216702
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,4096,3072,0.008206222620275285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,4096,3072,0.014098667436175875
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,3072,0.007932444413503012
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,4096,10240,0.022037333912319605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,4096,2560,0.013356444736321768
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,2560,0.140940441025628
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,8192,0.01311822235584259
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,4096,2048,0.0063928887248039246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,2560,0.0075902218619982404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,4096,7168,0.018925334016482036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,4096,1536,0.0053591111467944244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,4096,2048,0.013074666261672974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,2048,0.141466670566135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,2048,0.007288888924651676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,6144,0.011041777829329172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,4096,1024,0.004139555411206352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,4096,1536,0.013047110703256396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,1536,0.14271999730004206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,1536,0.006629333313968446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,5120,0.1459262238608466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,4096,1024,0.012075555821259817
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,1024,0.00629333323902554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,1024,0.14260355631510416
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,4096,512,0.0035546666218174826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,4096,768,0.012260444462299347
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,768,0.005984888722499211
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,768,0.14463644557529026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,4096,256,0.0032319999817344877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,4096,512,0.011854222251309289
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,512,0.005872000008821487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,512,0.14325333966149226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,4096,128,0.0028719999310043124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,4096,256,0.012065778175989786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,256,0.14692089292738172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,256,0.005906666732496685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,4096,64,0.002858666703104973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,4096,128,0.011742221812407175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,3072,0.13975377877553305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,4096,32,0.0027955555253558685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,4096,128,0.13679377237955728
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,4096,64,0.011695110963450538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3584,65536,0.07847377989027235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,4096,32,0.01168444421556261
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3584,16384,0.024488889508777197
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3584,65536,0.0857235524389479
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3584,16384,0.02834133307139079
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,65536,0.06718133555518256
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3584,12288,0.020094222492641874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,16384,0.14451289176940918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,16384,0.020277332928445604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,4096,768,0.0038373333712418876
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,65536,0.19455288516150582
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3584,12288,0.023642667465739783
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3584,10240,0.01789777808719211
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,12288,0.14407732751634386
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,12288,0.016742222838931613
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3584,8192,0.015285332997639975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,10240,0.014987554815080432
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,10240,0.14262044429779053
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3584,7168,0.013889777991506787
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3584,8192,0.019985778464211356
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,8192,0.01276000009642707
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,8192,0.14347555902269152
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3584,6144,0.013414222333166333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3584,7168,0.01904000010755327
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,7168,0.14115466011895075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,7168,0.011751111182901593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,4096,128,0.005947555518812603
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3584,5120,0.010844444235165914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3584,6144,0.01757600075668759
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,6144,0.010696888797812991
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,6144,0.14374844233194986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3584,4096,0.009685333404276106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3584,5120,0.016733333468437195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,5120,0.14314311080508763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,5120,0.009676444033781687
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3584,3584,0.008556444611814287
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3584,4096,0.015846222639083862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,4096,0.008370666868156856
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,4096,0.13973955313364664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3584,3072,0.00796977761718962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3584,3584,0.014466666513019137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,3584,0.008266666697131263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3584,2560,0.007303111255168915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3584,3072,0.014236445228258768
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,3072,0.0079644446571668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,3072,0.14758933915032282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3584,2048,0.006048000107208888
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3584,2560,0.013460444079505073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,2560,0.007650666766696506
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3584,2048,0.012789333032237159
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3584,10240,0.02125955621401469
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,2048,0.007231111327807109
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,2048,0.14301244417826334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3584,1024,0.0038942222793896994
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3584,1536,0.012786666552225748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,1536,0.006959999601046245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,1536,0.1392515500386556
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3584,1024,0.012441777520709567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3584,768,0.003543111185232798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,1024,0.006269333263238271
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,1024,0.14027110735575357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3584,512,0.0035457776652442086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3584,768,0.01239377756913503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,768,0.13981066809760198
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3584,512,0.012079111403889127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,512,0.0058853332367208265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,512,0.1424471139907837
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,3584,0.14095200432671443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3584,256,0.011690666278203329
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,256,0.14225155777401394
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3584,128,0.003129777808984121
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,256,0.00573600000805325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3584,128,0.011761777930789523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,128,0.13274489508734807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,3584,2560,0.1398435566160414
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,128,0.00591555568906996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3584,32,0.0028684443483750024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3584,1536,0.005005333158704969
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3584,64,0.011678222152921887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3072,65536,0.06951822174919976
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3072,65536,0.08388977580600315
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,4096,2560,0.007595555649863348
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3072,16384,0.02184533410602146
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,65536,0.06600622336069743
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,3584,768,0.006251555350091722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3072,16384,0.02791022260983785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,16384,0.02011022302839491
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,65536,0.1938906643125746
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,16384,0.14671911133660212
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3072,12288,0.018139556050300598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3072,10240,0.015303999185562134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3584,256,0.0031582222630580268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3072,12288,0.02269422180122799
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,12288,0.016499555773205228
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,12288,0.13962754938337538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3072,8192,0.013831111292044321
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3072,10240,0.02110044492615594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,10240,0.014431110686726041
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3072,7168,0.01309511148267322
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3072,8192,0.01943555474281311
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,8192,0.14657421906789145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,8192,0.012339555554919772
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3072,6144,0.011439111497667102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3072,7168,0.01828533411026001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3584,64,0.0028488888508743713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,7168,0.01125600023402108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,7168,0.14053955343034533
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3072,5120,0.010859555668301053
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3072,6144,0.017293334007263184
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,6144,0.010336000058386061
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3584,32,0.011778666741318174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,6144,0.14379999372694227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3072,4096,0.00963199966483646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3072,5120,0.0158897770775689
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,5120,0.14823110898335776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3072,3584,0.008576889004972246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,5120,0.009359111388524374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3072,4096,0.015572445260153877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,4096,0.14469422234429255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3072,3072,0.008039111064540016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3072,3584,0.014463999205165438
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,3584,0.008341333104504479
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3072,2560,0.006919999917348226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,3584,0.14019644260406494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3072,3072,0.01406133340464698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,3072,0.14278577433692083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3072,2560,0.013425778183672162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,2560,0.13893777794308132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,2560,0.007490666376219855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3072,2048,0.013081777426931592
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3072,1536,0.004561777744028303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,2048,0.14002754953172472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,10240,0.1402799950705634
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3072,1536,0.01275111072593265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,1536,0.006593777901596493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,1536,0.1391706731584337
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3072,1024,0.0042071110672420925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,1024,0.00637511122557852
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3072,768,0.003899555653333664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3072,768,0.012144889268610211
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,4096,0.008310221963458592
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,768,0.14542578326331243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,768,0.00591466658645206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3072,512,0.003463111196955045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,3072,0.00775022225247489
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3072,2048,0.006221333311663733
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3072,512,0.012042666474978128
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3072,256,0.00314666661951277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,512,0.1454319953918457
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3072,128,0.0028488888508743713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,256,0.005604444278611078
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,2048,0.007253333098358578
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3072,128,0.011749332977665795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3072,64,0.002554666664865282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,128,0.0058871110280354815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,3072,32,0.002809777855873108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3072,64,0.011376000112957425
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3072,32,0.011717333561844297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3072,1024,0.012504888905419244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2560,65536,0.05881600247489082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,1024,0.13932000266181097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2560,16384,0.019595555133289762
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2560,65536,0.08060889111624824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,65536,0.06507822540071276
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2560,16384,0.026130666335423786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2560,12288,0.016466667254765827
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,16384,0.019918221566412184
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,16384,0.1454275581571791
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2560,12288,0.022265776991844177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2560,10240,0.01479822231663598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,12288,0.015776889191733468
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,12288,0.14726132816738552
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,3072,256,0.011408888631396823
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2560,10240,0.02054933375782437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,10240,0.014264000786675347
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,10240,0.14649689197540283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2560,8192,0.012795555922720166
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2560,7168,0.01200355589389801
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,128,0.13317510816786024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,8192,0.012039110892348819
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2560,8192,0.01883555617597368
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,8192,0.14470399750603571
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2560,6144,0.011053333679835001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2560,7168,0.017937777770890128
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,7168,0.01069155583779017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,7168,0.14040088653564453
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2560,5120,0.010035555395815108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2560,6144,0.01735822194152408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,6144,0.010102222363154093
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,6144,0.14095466666751438
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2560,4096,0.008960888617568547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2560,5120,0.015448000695970325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,5120,0.009318222602208456
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,5120,0.1452817784415351
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2560,3584,0.008310221963458592
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2560,4096,0.014998222390810648
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,4096,0.140883551703559
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2560,3584,0.014089778065681458
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2560,3072,0.007507555186748505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,3584,0.008007110820876228
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2560,3072,0.01367555558681488
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,3072,0.14052266544765896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2560,2560,0.006625777731339137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,3072,0.007715555528799693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2560,2560,0.013229333692126803
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,2560,0.1449093288845486
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2560,2048,0.00574311117331187
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,2560,0.007303111255168915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2560,2048,0.013064000341627332
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,2048,0.14075644810994467
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2560,1536,0.004613333278232151
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,2048,0.0069360000391801195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2560,1536,0.012686221963829465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,4096,0.008332444561852349
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,1536,0.14473777347140843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2560,1024,0.0041759999261962045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,65536,0.19434132840898302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,1536,0.006623111251327727
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2560,768,0.0038497779104444715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2560,1024,0.012280888855457306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,1024,0.006245333287451003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,1024,0.13894932799869114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,3072,512,0.005922666854328579
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2560,512,0.0034764446318149567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2560,768,0.012100444071822695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,768,0.005945777727497949
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,768,0.14917688899570042
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2560,256,0.0031617778456873367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,3072,256,0.1425795555114746
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2560,512,0.011729777687125735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,512,0.13969243897332087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2560,128,0.0028497777465316984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2560,256,0.011577777564525604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,256,0.005586666779385672
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,256,0.14123822583092585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2560,64,0.002854222224818336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2560,128,0.011408888631396823
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2560,32,0.002833777831660377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,128,0.13732177681393093
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,128,0.005601777798599667
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2048,65536,0.047981331745783486
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2560,64,0.011699555648697747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2560,32,0.011745777395036487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2048,65536,0.07875555753707886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2048,16384,0.026113778352737427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,65536,0.193085339334276
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,16384,0.14833333757188585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2048,12288,0.015257777439223396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,16384,0.01941066649225023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2048,10240,0.014012444350454541
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2048,12288,0.0217884447839525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,2560,3584,0.14099821779463026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,12288,0.015240000353919135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,12288,0.14420089456770155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2048,8192,0.013099555340078143
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2048,10240,0.020271999968422782
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,10240,0.013753777576817406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,10240,0.1479840013715956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,8192,0.1446711089875963
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,2560,512,0.005893333504597346
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2048,7168,0.017553778158293832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2048,6144,0.011095110740926532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,7168,0.1446977721320258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,7168,0.010759111079904767
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2048,6144,0.016170667277442086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,6144,0.010000000397364298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2048,5120,0.009783999787436591
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2048,16384,0.017275555266274344
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2048,5120,0.015222221612930298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,5120,0.009325332939624786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,5120,0.1464808914396498
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,65536,0.06297422117657132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,4096,0.008339555727110969
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,4096,0.14848977989620632
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2048,3584,0.007524444825119442
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2048,3584,0.013786666923099093
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,3584,0.14348799652523464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2048,3072,0.006615110983451207
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2048,8192,0.01882133384545644
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,8192,0.01128088848458396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2048,3072,0.013594666288958656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2048,2560,0.006240889016124938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,3072,0.14383822017245823
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2048,2560,0.013424889081054263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,2560,0.14452621671888563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,6144,0.14451911714341906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2048,2048,0.0052391112678580815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,2560,0.007314666277832455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2048,2048,0.012789333032237159
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,2048,0.14422844515906438
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2048,1536,0.004874666531880696
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,2048,0.0069679998689227635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2048,4096,0.014431110686726041
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2048,1536,0.012758221891191272
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2048,1024,0.004026666697528628
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,1536,0.1440595520867242
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,1536,0.006568000134494569
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2048,1024,0.012444444828563266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,1024,0.006271111054552927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,1024,0.14899644586775038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2048,768,0.012071111136012606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,3072,0.007628444168302748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2048,512,0.0035111109415690103
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2048,512,0.011752888560295105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,512,0.14087733957502577
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2048,4096,0.008802666432327693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2048,256,0.0031511110977994073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,512,0.005873777800136142
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2048,256,0.011707555916574268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,256,0.1391297843721178
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2048,128,0.0028817777832349143
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,256,0.005579555614127054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2048,768,0.003516444315512975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,3584,0.008030222521887885
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2048,128,0.011738667057620155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,128,0.13329244984520808
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,128,0.00574311117331187
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2048,32,0.0028311111446883944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1536,65536,0.03786844346258376
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2048,64,0.011396444506115384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,2048,768,0.005899555567238066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1536,65536,0.07585510942671034
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1536,16384,0.015620445211728414
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,65536,0.19272532727983263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1536,16384,0.025359110699759588
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,16384,0.13976621627807617
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1536,12288,0.013217777841620974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,16384,0.018657777044508193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1536,12288,0.021915555000305176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,12288,0.14269777139027914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1536,10240,0.013159111142158508
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,12288,0.014120888378885059
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2048,7168,0.011666666302416058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,2048,64,0.0028799999919202593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1536,10240,0.020256888535287645
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1536,8192,0.012455110748608908
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,10240,0.1411715613471137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,10240,0.01278133359220293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,2048,768,0.14555733733707005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1536,7168,0.011057777537239922
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1536,8192,0.018208000395033095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,8192,0.14704533418019614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,8192,0.011347555451922946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1536,6144,0.010658666491508484
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1536,7168,0.01684977776474423
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,7168,0.010666666759385003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,7168,0.13954844739702013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1536,5120,0.009587555295891231
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,65536,0.062253336111704506
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1536,6144,0.01614488826857673
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,6144,0.14817867014143202
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,6144,0.010026666853162978
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1536,5120,0.01514400045077006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,5120,0.009296889106432596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,5120,0.14166133933597141
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1536,3584,0.007466666400432587
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1536,4096,0.014122666584120857
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,4096,0.14539910687340632
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1536,3072,0.0069360000391801195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,3584,0.007986666427718269
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1536,3584,0.013814222481515674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,3584,0.14289777808719212
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1536,2560,0.00590133335855272
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1536,3072,0.013415111435784234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,3072,0.1490159961912367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,3072,0.007651555869314406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1536,2048,0.005254222287072076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1536,2560,0.013396444420019785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,2560,0.007332444190979004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,2560,0.14152444733513728
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1536,1536,0.00452977791428566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1536,2048,0.013074666261672974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,2048,0.006953777538405524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1536,1024,0.003882666842805015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1536,1536,0.012426666915416718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,1536,0.006598222172922558
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,1536,0.1492720047632853
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1536,768,0.00350222239891688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1536,1024,0.012430222498046027
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,2048,32,0.012081777883900536
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,1024,0.14199911223517522
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1536,512,0.0035084444615576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1536,768,0.01204177737236023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,768,0.006239999913507038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,768,0.14361866315205893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1536,256,0.0028693332440323303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1536,512,0.01198933356338077
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1536,4096,0.009015111459626092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,512,0.005919111271699269
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,512,0.1462435589896308
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1536,128,0.002874666617976295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1536,256,0.011776888536082374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,256,0.14086843861473933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,256,0.005642666584915585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1536,64,0.0024986666523747975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,4096,0.00831822223133511
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1536,128,0.011422221859296164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1536,32,0.002806222273243798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,128,0.005622222191757626
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,128,0.13427110513051352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1024,65536,0.027319111757808264
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1536,64,0.011055111057228513
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1024,16384,0.01256711118751102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1024,65536,0.07147377729415894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,65536,0.05860622061623467
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1024,16384,0.02498666610982683
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,65536,0.193977779812283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,1536,2048,0.146978669696384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,16384,0.1449351045820448
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,16384,0.016857778032620747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1024,10240,0.010703110860453712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1024,12288,0.020448888341585796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,12288,0.014074666632546319
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,1536,1024,0.00627377794848548
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1024,10240,0.01910666624704997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1024,8192,0.010022222167915767
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,10240,0.14208266470167372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,10240,0.012710221939616732
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1024,7168,0.009647111097971598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1024,8192,0.01717866626050737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,8192,0.14147733317481145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1024,7168,0.016501333978441026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,7168,0.010596444209416708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1024,6144,0.008961777720186446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,7168,0.13926666312747532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,6144,0.010023111270533668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1024,6144,0.015826667348543804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,6144,0.14339289400312635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1536,32,0.011711111499203576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1024,5120,0.014919110470347934
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1024,4096,0.007930667036109501
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,5120,0.009324444664849175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1024,12288,0.012230222423871359
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1024,4096,0.014362666341993542
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1024,3584,0.0069546666410234236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,4096,0.15069954925113255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,4096,0.008345777789751688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,12288,0.14869688616858587
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1024,3584,0.014091556270917257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1024,3072,0.0069431112044387394
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,3584,0.008099555969238281
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,3584,0.14595377445220947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1024,2560,0.005664888769388199
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1024,3072,0.013436444103717804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,3072,0.007819555699825287
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,3072,0.14843111568027073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,8192,0.011309333145618439
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1024,2048,0.00527555536892679
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1024,2560,0.013073777986897362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,2560,0.007360888852013483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,2560,0.1424595514933268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1024,1536,0.00464088883664873
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1024,2048,0.013073777986897362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,2048,0.1481226682662964
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,2048,0.007110222346252865
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1024,1024,0.003882666842805015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1024,1536,0.012738666600651212
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,1536,0.14551022317674425
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1024,5120,0.008616000413894653
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1024,768,0.0038657776183552216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1024,1024,0.012058667010731168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,1024,0.006247111078765657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,1024,0.14823555946350098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1024,512,0.0035048888789282907
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1024,768,0.012088889049159156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,768,0.005956444475385878
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1024,512,0.011439999772442712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,512,0.14254489209916857
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1024,256,0.011396444506115384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,256,0.14139466815524632
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1024,128,0.0028719999310043124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,256,0.005591111050711737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1024,128,0.011732444167137146
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,128,0.13872622119055852
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1024,64,0.0028026666906144884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,128,0.005601777798599667
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1024,32,0.00286311118139161
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1024,64,0.011057777537239922
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,768,65536,0.024703999360402424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,1024,32,0.011708444191349877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,1536,0.006637333167923822
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,768,65536,0.07018755541907416
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,65536,0.058317330148484975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,65536,0.19188533888922799
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,5120,0.14231733481089273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,768,16384,0.023950222465727065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,1024,768,0.14589333534240723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,768,12288,0.009993777506881291
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,1024,256,0.002874666617976295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,16384,0.016895999511082966
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,16384,0.1422951089011298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,1024,512,0.005872000008821487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,768,10240,0.010335110955768162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,768,12288,0.01992266707950168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,12288,0.014058666096793281
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,768,8192,0.009376000199053023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,10240,0.012784000072214337
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,10240,0.14134932888878718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,768,7168,0.00889866633547677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,768,8192,0.01686044368478987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,8192,0.011171555353535546
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,8192,0.14754577477773032
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,768,6144,0.007992888490358988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,768,7168,0.016515556308958266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,7168,0.01070577816830741
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,7168,0.1454426712459988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,768,5120,0.008349333372380998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,768,6144,0.015541333291265698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,6144,0.010017777482668558
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,6144,0.13982755608028835
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,768,4096,0.007310222420427535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,768,5120,0.01479822231663598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,5120,0.14945156044430202
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,5120,0.009306666751702627
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,768,3584,0.0069724445541699724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,768,4096,0.014422222971916199
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,768,16384,0.011325333681371478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,4096,0.14384622044033474
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,768,3072,0.006567111031876669
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,768,3584,0.013816888961527081
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,3584,0.007972444097201029
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,3584,0.13999911149342856
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,768,2560,0.005607111172543631
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,768,3072,0.013423999978436364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,3072,0.14028088251749674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,3072,0.007648000286685095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,768,2048,0.005048000150256687
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,12288,0.14602133962843153
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,2560,0.007313778003056844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,2560,0.14445688989427355
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,768,1536,0.004510222209824456
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,768,2048,0.013060444758998023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,2048,0.1438515583674113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,768,1536,0.012436444560686747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,768,1024,0.003834666477309333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,1536,0.14339378145005968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,1536,0.0068408888247278
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,768,1024,0.012103111379676394
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,1024,0.14252889156341553
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,768,768,0.0035537779331207275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,768,768,0.012101333174440594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,768,0.1453591055340237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,4096,0.008319999608728621
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,768,512,0.003216000066863166
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,768,0.005930666708283954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,768,512,0.011405333048767514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,512,0.1450079944398668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,768,256,0.0028728888266616394
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,512,0.005897777775923411
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,256,0.005589333259397083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,768,2560,0.01312444441848331
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,256,0.1482586728201972
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,768,128,0.0028159999185138275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,768,10240,0.0185244447655148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,768,64,0.0025840000145965153
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,768,128,0.011734222372372946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,2048,0.007257777783605788
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,768,32,0.002805333377586471
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,768,64,0.011529777612951068
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,512,65536,0.021511110994550917
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,768,32,0.011742221812407175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,512,65536,0.06820444266001384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,1024,0.006288888967699475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,65536,0.19196800390879312
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,512,16384,0.011064888702498542
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,512,16384,0.02373955481582218
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,16384,0.01682222220632765
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,512,12288,0.008988444175985124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,16384,0.13972711563110352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,12288,0.0140657772620519
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,512,12288,0.01994311147265964
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,12288,0.14536355601416692
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,512,10240,0.009323555562231276
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,10240,0.012732444538010491
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,10240,0.14436622460683188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,512,8192,0.008323555191357931
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,768,256,0.011516444385051727
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,512,8192,0.01687111126052009
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,8192,0.13894489076402453
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,512,7168,0.007962666451931
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,768,128,0.1391999986436632
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,512,7168,0.016520889268981088
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,7168,0.010523555179437002
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,7168,0.1480835543738471
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,512,6144,0.008674666285514832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,512,6144,0.015245333313941956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,65536,0.05713688664966159
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,512,5120,0.007968000239796108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,6144,0.14345244566599527
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,512,5120,0.014505777094099255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,5120,0.009349333743254343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,5120,0.13910667101542154
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,512,4096,0.007287110719415877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,512,4096,0.014404444230927361
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,512,3584,0.006950221955776215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,512,10240,0.018527110417683918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,512,3584,0.014132445057233175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,3584,0.007965332931942409
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,3584,0.14102578163146973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,8192,0.011063999599880643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,512,3072,0.007057777709431118
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,768,128,0.005624889085690181
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,512,2560,0.005586666779385672
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,512,3072,0.013437333206335703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,3072,0.007640889121426477
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,3072,0.1461822191874186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,512,2048,0.005224888937340842
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,2560,0.007329777710967594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,2560,0.14179821809132895
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,6144,0.009983110758993361
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,512,2048,0.01278133359220293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,512,1536,0.004518222063779831
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,2048,0.006980444822046492
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,2048,0.1442942221959432
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,512,1024,0.0038577777643998465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,512,1536,0.012446222205956778
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,1536,0.006619555668698416
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,1536,0.14074933528900146
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,512,768,0.00350577798154619
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,512,1024,0.012422222230169507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,4096,0.00832088871134652
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,1024,0.006249777972698212
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,1024,0.14373421669006348
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,512,512,0.0034453332838084963
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,512,768,0.012076444096035428
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,768,0.005934222290913264
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,768,0.14507467216915554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,512,256,0.002842666581273079
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,512,512,0.011772444678677453
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,512,0.1419235600365533
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,512,0.00592533333433999
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,512,128,0.00282577777074443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,512,256,0.011748444702890186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,256,0.005555555638339784
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,256,0.14767377906375462
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,512,64,0.002522666628162066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,512,128,0.011702222128709158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,512,32,0.0028239999794297745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,512,128,0.005607999861240387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,128,0.13316533300611708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,256,65536,0.015445333388116626
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,512,64,0.01071999967098236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,256,16384,0.009726222190592024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,512,32,0.01146577795346578
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,512,2560,0.013312000367376538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,256,65536,0.0664186676343282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,65536,0.19137688477834067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,256,16384,0.022973333795865376
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,16384,0.01687022215790219
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,16384,0.14073067241244847
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,256,10240,0.008973333570692275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,256,12288,0.019903999235894945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,12288,0.014106666048367819
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,12288,0.1475528875986735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,256,8192,0.008354666332403818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,256,10240,0.01863911085658603
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,10240,0.01276266657643848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,10240,0.14098400539822048
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,256,7168,0.00798044436507755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,256,8192,0.016494222813182406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,512,4096,0.13994044727749294
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,8192,0.1416106621424357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,256,6144,0.007297777467303806
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,256,7168,0.016536888149049547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,7168,0.14137689272562662
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,256,5120,0.006968888971540663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,256,6144,0.015448000695970325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,6144,0.00998933364947637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,6144,0.1411315600077311
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,256,4096,0.006617777877383762
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,256,5120,0.01499644418557485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,5120,0.009313777916961247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,5120,0.13883733749389648
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,256,3584,0.006223111102978389
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,256,4096,0.014092443717850579
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,4096,0.008298666940795051
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,4096,0.1423342227935791
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,256,3072,0.006592000110281839
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,256,3584,0.013761777844693927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,65536,0.05071022113164266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,3584,0.007994666695594788
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,256,12288,0.009677333136399588
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,256,2560,0.006280888699822956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,256,3072,0.01370844410525428
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,3072,0.007612444460391998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,3072,0.1395448843638102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,256,2048,0.006075555665625467
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,256,2560,0.013439111411571503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,2560,0.1443715492884318
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,256,2048,0.012759110993809171
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,2048,0.14575733078850642
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,2048,0.0069724445541699724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,8192,0.01128799964984258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,256,1536,0.012438221938080259
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,256,1024,0.0038515557017591265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,1536,0.14235733615027532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,1536,0.006615110983451207
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,256,1024,0.012097777591811286
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,1024,0.006269333263238271
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,1024,0.1430479950375027
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,256,768,0.0035777779089079965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,256,768,0.012085333466529846
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,768,0.005947555518812603
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,256,512,0.00314666661951277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,256,512,0.011754666765530905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,3584,0.14225244522094727
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,512,0.13974489106072321
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,256,256,0.0028773333049482773
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,512,0.005906666732496685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,256,256,0.01181422256761127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,256,128,0.002509333400262727
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,256,0.005571555760171678
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,2560,0.007291555404663086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,256,1536,0.004221333397759331
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,128,0.005608888963858287
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,256,64,0.0028320000403457214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,128,0.1361831161710951
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,256,32,0.002480000050531493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,128,65536,0.015087111128701104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,256,64,0.011415999796655444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,256,32,0.011707555916574268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,256,7168,0.010363555616802638
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,128,16384,0.00870577742656072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,128,65536,0.06478489107555814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,65536,0.050276445017920605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,65536,0.1914800008138021
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,128,12288,0.0076453329788313965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,128,16384,0.02276355524857839
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,16384,0.13907821973164877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,16384,0.01682400041156345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,128,10240,0.007242666350470648
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,128,12288,0.01994311147265964
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,12288,0.1479991144604153
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,12288,0.014110222458839417
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,128,8192,0.0069022224181228215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,128,10240,0.018416888184017606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,10240,0.14153956042395696
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,128,8192,0.016857778032620747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,128,7168,0.006605333338181178
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,8192,0.01107377807299296
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,8192,0.14636622534857854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,256,0.14147555828094482
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,128,6144,0.00628266649113761
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,128,7168,0.01626044511795044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,7168,0.010458666417333813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,256,128,0.011707555916574268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,7168,0.1412880023320516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,128,5120,0.006162666612201267
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,128,6144,0.015519110692871941
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,6144,0.009988444546858469
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,128,4096,0.006566222343179915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,5120,0.009206222163306342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,128,5120,0.014800000521871777
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,5120,0.1400231122970581
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,128,3584,0.006257777826653586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,128,4096,0.014129777749379476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,4096,0.14241688781314427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,128,3072,0.005903999838564131
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,128,3584,0.013754666679435305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,3584,0.1413280036714342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,3584,0.007977777885066139
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,128,2560,0.006260444306664997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,128,3072,0.013413333230548434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,3072,0.007652444144090016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,256,768,0.14237688647376165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,128,2560,0.01313066648112403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,128,2048,0.006106666806671355
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,2560,0.007295110987292395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,2560,0.14227378368377686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,128,1536,0.004232888834344016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,128,2048,0.013052444491121503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,2048,0.14146488242679173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,2048,0.0069102222720781965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,128,1024,0.0038933331767717996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,128,1536,0.012443555725945367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,1536,0.006577777779764599
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,1536,0.14018932978312174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,128,768,0.003546666767862108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,128,1024,0.012087999946541257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,1024,0.006239999913507038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,1024,0.1421360042360094
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,128,512,0.0031439999325407874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,128,768,0.012090666426552666
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,768,0.005952000204059813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,768,0.13891555203331843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,6144,0.14481955104404026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,128,512,0.011725333001878528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,512,0.14004000027974448
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,512,0.005884444548024072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,4096,0.008282666405042013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,128,256,0.011759111450778114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,128,128,0.0028195555011431375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,256,0.14036088519626194
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,256,0.005599110904667113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,128,64,0.0028008888992998335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,128,128,0.011924444801277585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,128,0.00562755556570159
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,128,32,0.0024746666765875286
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,128,0.14150222142537436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,48,128,3072,0.1404071119096544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,64,65536,0.012171555724408893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,128,64,0.011701333026091257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,64,16384,0.00793422261873881
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,64,12288,0.007605333295133378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,128,32,0.011682666838169098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,64,10240,0.0070293330483966405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,64,8192,0.006970666348934174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,64,65536,0.06241244739956326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,64,12288,0.01992177797688378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,64,7168,0.00658044425977601
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,64,10240,0.018202667435010273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,64,6144,0.006296000132958095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,64,8192,0.017179555363125272
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,64,5120,0.006230222268237009
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,64,7168,0.016487111647923786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,64,4096,0.005576888720194499
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,64,6144,0.015477332803938123
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,64,3584,0.006258666515350342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,64,3072,0.0059262220230367445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,64,5120,0.015122666954994202
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,64,2560,0.006291555447710886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,64,4096,0.014418666561444601
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,64,3072,0.013395555317401886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,64,2560,0.013168888787428537
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,64,1536,0.004350222233268949
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,64,2048,0.012740444805887012
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,64,1024,0.0038453332251972626
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,64,1536,0.012723555167516073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,64,768,0.003530666646030214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,64,1024,0.012091555529170565
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,64,512,0.00314666661951277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,64,768,0.012094222009181976
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,64,256,0.0028195555011431375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,64,512,0.011711111499203576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,64,128,0.0028026666906144884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,48,128,10240,0.012744888663291931
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,64,64,0.0024906665914588505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,64,256,0.011781333221329583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,64,32,0.0026364444444576898
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,64,128,0.011398221883508893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,64,64,0.01071911139620675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,64,32,0.011744000017642975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,32,16384,0.007939555578761632
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,32,65536,0.06188266807132297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,64,3584,0.013800889253616333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,32,12288,0.007295110987292395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,64,2048,0.006222222414281633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,32,10240,0.006966222491529252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,32,8192,0.006274666637182236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,32,12288,0.019670221540662978
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,32,10240,0.0185199992524253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,32,8192,0.01651822196112739
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,32,6144,0.006100444330109491
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,32,7168,0.01648355523745219
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,32,5120,0.006227555374304454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,32,6144,0.01520088811715444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,32,4096,0.005933333188295364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,32,5120,0.014816888504558139
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,64,16384,0.023355555203225877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,32,3584,0.006237333433495627
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,32,3072,0.005894222193294101
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,32,3584,0.014114666316244336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,32,3072,0.013760888742076026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,32,65536,0.012033777932325998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,32,2048,0.005613333235184352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,32,2560,0.013396444420019785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,32,1536,0.004505777938498391
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,32,2048,0.012728888955381183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,32,1024,0.003906666818592283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,32,16384,0.02331111166212294
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,32,768,0.003556444413132138
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,32,1536,0.012728888955381183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,32,512,0.0031555555760860443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,32,1024,0.01205599970287747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,32,7168,0.006599999964237213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,32,256,0.0028560000161329904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,32,768,0.012105777859687805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,32,512,0.011899555722872416
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,32,64,0.002530666689078013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,32,256,0.011696000066068439
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,32,128,0.011704000333944956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,32,32,0.0025315555847353404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,32,64,0.011545778148704104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,32,32,0.011710222396585675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,48,32,4096,0.013783111340469785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,65536,16384,0.17504089408450654
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,65536,16384,0.3638915485805935
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,32,2560,0.006259555617968242
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,65536,12288,0.2728373209635417
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,65536,12288,0.1325226624806722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,16384,0.13811111450195312
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,128,256,0.0027991111079851785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,65536,10240,0.22776622242397734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,65536,10240,0.11334310637580024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,48,32,128,0.0028195555011431375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,12288,0.10539644294314915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,65536,8192,0.18317333857218424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,65536,8192,0.09750044345855713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,10240,0.08932533529069688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,65536,7168,0.1620533333884345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,65536,7168,0.08334755235248142
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,8192,0.0718053314420912
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,12288,0.24167911211649576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,65536,6144,0.13812355200449625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,16384,0.3002435631222195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,10240,0.20782932970258924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,7168,0.06487022505866157
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,8192,0.17792710993025038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,65536,5120,0.11666044923994277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,65536,6144,0.07419200075997247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,6144,0.056236445903778076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,65536,4096,0.0957884457376268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,7168,0.16299200057983398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,65536,5120,0.06431733237372504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,5120,0.052566223674350314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,65536,3584,0.08327377504772611
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,65536,4096,0.05451733205053541
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,6144,0.14785333474477133
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,4096,0.040095110734303795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,5120,0.14724000295003256
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,65536,3072,0.07152711020575629
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,3584,0.03561066587766012
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,4096,0.14261689451005724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,65536,3072,0.04708266589376661
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,65536,2560,0.06008533636728922
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,3072,0.03165244393878513
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,3584,0.14273422294192845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,65536,2048,0.048660443888770215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,65536,2560,0.04189244574970669
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,2560,0.028074665202034846
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,3072,0.14798755115932888
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,65536,1536,0.03765155540572272
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,2048,0.02331999937693278
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,2560,0.14617955684661865
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,65536,1536,0.030696001317765977
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,65536,1024,0.02660888930161794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,2048,0.14367022779252794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,1536,0.01903288894229465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,1536,0.14783555931515163
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,65536,1024,0.024069334069887798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,1024,0.015127110812399121
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,1024,0.14531377951304117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,65536,512,0.014288889037238227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,65536,768,0.020590222544140287
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,768,0.14696978198157415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,768,0.013394667042626275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,65536,256,0.008336000144481659
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,65536,512,0.018253333038753934
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,512,0.011693333586057028
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,512,0.14121245013342962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,65536,128,0.006907555378145642
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,65536,256,0.015790222419632804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,256,0.0107022225856781
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,256,0.14836177561018202
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,65536,64,0.0059279998143513995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,65536,32,0.005929777605666055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,65536,128,0.13412266307406956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,65536,128,0.014379555980364481
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,65536,64,0.015505777464972602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,65536,32,0.01651733285850949
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,65536,2048,0.03686844309171041
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,16384,65536,0.35666844579908585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,16384,16384,0.07974221971299914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,65536,768,0.02056533263789283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,16384,16384,0.05810666746563382
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,65536,0.16436178154415554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,16384,12288,0.06182577874925402
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,16384,0.04408177733421326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,16384,12288,0.04614666766590542
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,65536,3584,0.053895112540986806
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,12288,0.03432088759210374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,65536,128,0.01072355525361167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,16384,10240,0.04105510976579454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,16384,8192,0.043824000491036304
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,10240,0.03086844417783949
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,10240,0.14395377371046278
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,16384,8192,0.03661155700683594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,16384,65536,0.19988089137607148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,16384,7168,0.03750755720668369
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,8192,0.025204444924990337
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,16384,7168,0.03246577911906772
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,8192,0.14396088653140598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,16384,6144,0.03276800115903219
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,7168,0.022527111901177302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,7168,0.14256800545586482
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,16384,6144,0.030060443613264296
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,16384,5120,0.028413332170910303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,6144,0.02034311162100898
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,6144,0.143103109465705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,16384,5120,0.02693333393997616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,16384,0.13963377475738525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,16384,4096,0.02401955591307746
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,5120,0.14062400658925375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,5120,0.01984177695380317
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,16384,3584,0.020672000116772123
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,16384,4096,0.023050665855407715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,12288,0.1428826649983724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,4096,0.015549333559142219
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,4096,0.1447395616107517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,16384,3072,0.0184542222155465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,16384,3584,0.023990222149425085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,3584,0.014526221487257214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,3584,0.14161954985724554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,16384,3072,0.021988444858127173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,3072,0.14042666223314074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,3072,0.013417777915795645
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,16384,2048,0.013152889079517789
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,16384,2560,0.020616888999938965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,2560,0.1464622285630968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,65536,0.23417867554558647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,16384,2048,0.01884355478816562
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,2048,0.13993422190348306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,2048,0.011047999891969891
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,16384,1536,0.010783111055692038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,16384,1024,0.008401778009202745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,16384,1536,0.017143110434214275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,1536,0.009700444837411245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,16384,1024,0.01445777714252472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,16384,768,0.006924444602595435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,1024,0.007961778177155389
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,16384,10240,0.05150933398140801
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,768,0.14436622460683188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,16384,768,0.013777777552604675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,768,0.007963555554548899
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,16384,512,0.005453333258628845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,16384,512,0.012738666600651212
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,512,0.0069324444565508105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,16384,256,0.003522666792074839
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,512,0.1438773340649075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,16384,2560,0.015818667080667283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,16384,128,0.003141333245568805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,256,0.006616000086069107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,16384,256,0.012216000093354119
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,256,0.14756088786655003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,16384,64,0.0031431110368834603
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,16384,128,0.012044444680213928
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,128,0.006632888896597757
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,128,0.13458222813076443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,16384,64,0.011496889094511667
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,16384,32,0.012416000167528788
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,1536,0.13892800278133816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,12288,65536,0.2566915618048774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,12288,65536,0.16677244504292807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,16384,1024,0.14038311110602483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,12288,16384,0.0729422238137987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,12288,16384,0.05148799882994758
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,16384,0.039592888620164655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,12288,12288,0.05749066670735677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,65536,0.14498133129543728
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,16384,0.1406631072362264
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,12288,10240,0.047409776184293956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,12288,12288,0.04124800033039517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,12288,0.032560888263914324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,12288,0.14698310693105063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,16384,2560,0.0123857781291008
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,16384,32,0.0028595555987623003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,10240,0.026752889156341553
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,10240,0.14349155955844456
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,12288,8192,0.033194667763180204
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,8192,0.022310222188631695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,12288,7168,0.03319022059440613
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,65536,0.20148089196946886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,12288,7168,0.02955555584695604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,7168,0.020455110404226515
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,12288,6144,0.02879555688963996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,7168,0.141485333442688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,12288,6144,0.027080888549486797
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,12288,5120,0.024633778466118708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,6144,0.018625777628686693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,6144,0.1444000005722046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,12288,5120,0.023919110496838886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,5120,0.017097777790493436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,5120,0.13852266470591226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,12288,3584,0.018239999810854595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,12288,4096,0.02166044380929735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,4096,0.014479999740918478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,4096,0.14239911238352457
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,12288,3072,0.016519111063745286
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,12288,3584,0.02181333303451538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,3584,0.14461867014567056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,12288,3072,0.020576889316240948
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,12288,10240,0.03687999977005853
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,3072,0.13973600334591335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,12288,2560,0.013836444252067141
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,3072,0.01274577776590983
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,12288,2048,0.011688888900809817
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,2560,0.011573333707120685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,2560,0.13800266053941515
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,12288,2048,0.017817777064111497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,2048,0.1420604387919108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,2048,0.010000000397364298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,12288,1536,0.00943911075592041
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,12288,1536,0.016286222471131217
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,1536,0.14266133308410645
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,12288,4096,0.021070222059885662
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,12288,1024,0.007415999968846639
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,1536,0.008988444175985124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,12288,1024,0.01368977791733212
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,1024,0.1410577827029758
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,12288,768,0.006554666492674086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,1024,0.007917332980367873
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,12288,768,0.013089777694808112
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,3584,0.013501333693663279
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,768,0.007623111208279927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,12288,512,0.005222222043408289
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,768,0.14321422576904297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,12288,256,0.003547555456558863
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,12288,8192,0.03793688946300083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,12288,512,0.012778666284349231
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,512,0.006867555694447623
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,512,0.143303116162618
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,12288,128,0.0035217776894569397
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,12288,256,0.012157333393891653
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,12288,2560,0.019242667489581637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,256,0.14600443840026855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,12288,64,0.0031306667046414483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,12288,128,0.012079999678664737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,8192,0.1400008863872952
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,12288,32,0.003142222141226133
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,12288,128,0.13388710551791722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,128,0.006591111007663939
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,12288,64,0.0110880004035102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,12288,32,0.012111110819710625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,10240,16384,0.06195022000206841
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,10240,65536,0.2320915593041314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,10240,65536,0.16296177440219456
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,10240,16384,0.04980088935958015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,10240,12288,0.04795022143257988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,65536,0.14062310589684382
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,10240,12288,0.040135112073686384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,16384,0.03792355457941691
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,16384,0.14705599678887263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,10240,10240,0.040998223755094744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,12288,0.0318853325313992
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,12288,0.13990044593811035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,10240,8192,0.035439110464519925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,10240,10240,0.03590577840805054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,10240,0.026120889517996047
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,10240,0.1456835534837511
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,10240,7168,0.02782577938503689
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,8192,0.021527111530303955
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,10240,8192,0.03176711002985636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,8192,0.14316888650258383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,10240,6144,0.024302222662501868
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,12288,256,0.0063973334100511335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,65536,0.1893528832329644
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,7168,0.019659555620617337
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,7168,0.14535378085242376
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,10240,5120,0.02049688829316033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,10240,6144,0.02530844509601593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,10240,4096,0.0173662222094006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,6144,0.14116888576083714
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,10240,5120,0.023247111174795363
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,5120,0.016921778519948322
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,5120,0.14510044786665174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,10240,3584,0.015455999308162265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,4096,0.01407911048995124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,10240,4096,0.020972443951500788
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,4096,0.14083733823564318
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,10240,3072,0.013419555293189155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,10240,3584,0.02092444399992625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,3584,0.013385777672131857
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,3584,0.1420382261276245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,10240,3072,0.01975022256374359
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,3072,0.012102222277058495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,10240,2560,0.018194667167133756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,10240,2048,0.010973333484596677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,2560,0.1406666702694363
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,2560,0.010776888993051318
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,10240,2048,0.017165333032608032
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,2048,0.009562667045328353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,2048,0.1403848859998915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,10240,1536,0.015503111812803479
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,10240,7168,0.028464890188641016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,1536,0.139847993850708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,10240,1024,0.007642666498819987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,1536,0.008676444490750631
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,6144,0.018008889423476327
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,1024,0.14841155211130777
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,1024,0.007641777396202087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,10240,768,0.013063111239009433
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,768,0.007592000067234039
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,768,0.14724355273776585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,10240,512,0.004767111192146937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,10240,2560,0.012063999970753988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,512,0.14584444628821477
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,10240,256,0.011390222443474663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,10240,1536,0.008964444200197855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,256,0.14538578192392984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,256,0.006276444428496891
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,10240,128,0.003544888976547453
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,10240,1024,0.013750221994188098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,10240,768,0.005997333261701796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,10240,128,0.011395555403497485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,10240,64,0.0031840000301599503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,128,0.13227821720971003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,10240,32,0.00286666676402092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,10240,64,0.011053333679835001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,10240,512,0.012734221915404001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,10240,3072,0.14545067151387533
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,8192,65536,0.17678933673434785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,8192,65536,0.11211555533938938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,8192,16384,0.047240889734692044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,65536,0.08052355713314481
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,8192,16384,0.0376800000667572
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,16384,0.024625778198242188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,8192,12288,0.03640977872742547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,65536,0.18570755587683785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,128,0.00657155571712388
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,8192,12288,0.030886222918828327
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,8192,10240,0.02797422144148085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,10240,512,0.006607111129495833
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,10240,256,0.003569777641031477
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,12288,0.01975555552376641
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,8192,10240,0.028111110130945843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,8192,8192,0.023959111836221483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,12288,0.24520267380608451
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,10240,0.01742488808102078
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,10240,0.13938221666547987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,8192,7168,0.02089866664674547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,8192,8192,0.025281776984532673
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,8192,7168,0.022202667262819078
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,8192,6144,0.018311111463440787
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,7168,0.14149955908457437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,8192,0.2287635538313124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,8192,6144,0.020764445265134174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,6144,0.013235555754767524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,6144,0.1409297784169515
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,16384,0.147873772515191
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,8192,5120,0.019239111079110038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,5120,0.14106310738457575
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,5120,0.012411555482281579
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,8192,4096,0.017880888448821176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,10240,32,0.01202311118443807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,8192,3584,0.01241422196229299
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,4096,0.14681866433885363
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,4096,0.010664888554149203
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,8192,3584,0.018292443619834054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,3584,0.009693333672152625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,3584,0.14749599827660456
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,8192,3072,0.011026666396194033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,3072,0.009339555270142024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,8192,3072,0.017492444978819955
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,3072,0.14701244566175672
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,8192,2560,0.00963288876745436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,7168,0.013811555173661975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,8192,0.015040000279744467
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,8192,2048,0.008575110799736446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,8192,2560,0.01647733317481147
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,2560,0.008851555486520132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,2560,0.1458079947365655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,8192,2048,0.015114666687117683
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,2048,0.1457644436094496
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,2048,0.00794400026400884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,8192,1024,0.005547555370463266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,8192,1536,0.01384622189733717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,1536,0.007279111279381647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,8192,4096,0.013707555830478668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,8192,1024,0.01276533305644989
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,8192,768,0.004543999830881755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,1024,0.006616888774765863
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,1024,0.1458924478954739
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,8192,512,0.0038684445122877755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,8192,768,0.012402666939629449
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,768,0.006256888724035687
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,768,0.1405680047141181
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,8192,256,0.0031715554909573663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,8192,512,0.011801777614487542
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,512,0.14359288745456272
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,512,0.006065777606434292
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,8192,128,0.0031786666562159858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,8192,256,0.011730666789743634
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,256,0.1420631143781874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,8192,64,0.0028213332924577924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,8192,128,0.011712888876597086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,128,0.005945777727497949
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,128,0.1350568930308024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,8192,32,0.002926222152180142
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,8192,64,0.011747555600272285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,8192,32,0.011712888876597086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,7168,65536,0.1530862251917521
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,7168,65536,0.10611199670367771
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,7168,16384,0.04394488864474826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,8192,1536,0.0076142218377855085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,8192,5120,0.01606399979856279
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,7168,16384,0.035616000493367515
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,7168,12288,0.03483555714289347
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,16384,0.023002665903833177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,16384,0.14521422651078966
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,7168,12288,0.029106666644414265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,7168,10240,0.029783109823862713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,12288,0.14000266128116184
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,12288,0.018566222654448617
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,65536,0.18464088439941406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,10240,0.016553777787420485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,10240,0.13728177547454834
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,7168,8192,0.022317333353890315
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,7168,8192,0.022687110635969374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,8192,0.014558222558763293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,8192,0.1387537717819214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,7168,7168,0.018903111418088276
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,7168,7168,0.021278222401936848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,7168,0.013529777526855469
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,7168,6144,0.016562667157914903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,7168,0.13781600528293186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,7168,6144,0.020187555087937247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,6144,0.012549333274364471
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,65536,0.07508622275458442
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,7168,5120,0.014595554934607612
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,6144,0.14146578311920166
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,8192,1536,0.14327111509111193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,7168,4096,0.012757333616415659
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,7168,5120,0.018959111637539335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,5120,0.011853333148691388
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,5120,0.14236977365281847
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,7168,3584,0.011130666567219628
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,7168,4096,0.01750577820671929
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,4096,0.1420933273103502
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,4096,0.009691555466916826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,7168,3072,0.009995555712117089
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,7168,3584,0.017831999394628737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,3584,0.009163555171754625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,7168,3072,0.017059556312031217
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,7168,2560,0.009288888838556077
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,3072,0.008628444539176093
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,7168,10240,0.026385777526431616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,7168,2048,0.007799111306667328
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,2560,0.007947555846638149
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,2560,0.29843910535176593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,2048,0.1399386723836263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,7168,2048,0.014505777094099255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,7168,1536,0.007248000138335758
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,2048,0.00741244438621733
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,1536,0.13998488585154215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,7168,1536,0.013569778038395775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,7168,1024,0.0052533331844541765
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,1536,0.0070062221752272705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,1024,0.1392399999830458
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,7168,1024,0.012841777669058906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,7168,768,0.004576000074545543
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,1024,0.006441777778996362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,3584,0.14528977870941162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,768,0.13980178038279215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,7168,512,0.003527111063400904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,7168,768,0.012510221865442065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,3072,0.14194666014777288
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,8192,256,0.005600000007285013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,7168,256,0.00315022220214208
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,768,0.00629066675901413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,7168,2560,0.015948444604873657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,512,0.14565689033932155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,7168,256,0.011400000088744693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,512,0.006016888966162999
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,256,0.1405057774649726
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,7168,64,0.0028195555011431375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,256,0.00599644457300504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,7168,128,0.13662577999962702
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,7168,128,0.0058311112225055695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,7168,64,0.011769777370823754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,7168,32,0.011041777829329172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,6144,65536,0.1249119970533583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,6144,65536,0.09555821948581272
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,6144,16384,0.03755111164516873
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,65536,0.07247644662857056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,6144,16384,0.03328088919321696
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,16384,0.14377244313557944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,7168,512,0.012373333175977072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,16384,0.022036444809701707
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,65536,0.18393777476416695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,7168,128,0.003132444495956103
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,6144,12288,0.027676444914605882
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,12288,0.018011555075645447
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,12288,0.1412364509370592
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,6144,10240,0.023524444964196947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,7168,32,0.0028115556471877625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,10240,0.015843555331230164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,10240,0.14035111003451878
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,6144,7168,0.01980533368057675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,6144,8192,0.02198666665289137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,8192,0.14356177382998997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,6144,7168,0.020581333173645865
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,6144,6144,0.01754133403301239
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,7168,0.1394719945059882
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,7168,0.013191999660597907
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,6144,6144,0.019290667441156175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,6144,0.012019555601808759
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,6144,0.13990844620598689
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,6144,5120,0.018184888694021437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,5120,0.13880088594224718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,6144,12288,0.03039466672473484
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,6144,4096,0.011611555185582904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,5120,0.011378666592968835
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,6144,10240,0.026047110557556152
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,7168,128,0.011738667057620155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,6144,3584,0.010450666977299584
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,6144,8192,0.022322666313913133
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,4096,0.009429333110650381
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,4096,0.13908355765872532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,6144,4096,0.0170462214284473
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,6144,3072,0.009279111193286048
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,6144,3584,0.01716800034046173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,3584,0.00888000014755461
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,3584,0.14116177293989393
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,6144,2560,0.008590222232871586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,8192,0.014044443766276041
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,3072,0.008014221986134848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,3072,0.13933777809143066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,6144,2048,0.0074622225430276655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,6144,2560,0.015012444721327888
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,2560,0.007442666424645319
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,2560,0.14059200551774767
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,6144,1536,0.006923555499977536
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,2048,0.007304888632562425
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,2048,0.14402843846215144
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,6144,1024,0.0052897776994440295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,6144,1536,0.013396444420019785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,6144,5120,0.015493333339691162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,1536,0.006928000185224745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,6144,768,0.004345777961942884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,6144,1024,0.012747555143303342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,1024,0.006239111224810283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,1024,0.14057600498199463
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,6144,512,0.003530666646030214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,6144,768,0.012109333442317115
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,768,0.006263999889294307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,768,0.1468693282869127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,6144,256,0.0031840000301599503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,6144,512,0.0120319997270902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,512,0.005886222339338726
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,512,0.1426408953136868
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,6144,128,0.0028693332440323303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,6144,256,0.011721777419249216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,256,0.13925066259172228
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,256,0.005896889087226655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,6144,64,0.0026764445420768526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,6144,128,0.011365333365069495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,6144,32,0.0028204443968004654
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,6144,128,0.00591822216908137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,6144,64,0.011409777734014722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,6144,3072,0.016570667425791424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,5120,65536,0.1095493303404914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,5120,65536,0.08769866493013169
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,6144,2048,0.014127999544143677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,5120,16384,0.032935112714767456
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,65536,0.0700168874528673
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,1536,0.14609955416785345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,5120,16384,0.031121777163611516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,5120,12288,0.026743110683229234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,16384,0.021282666259341772
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,16384,0.14063199361165366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,5120,12288,0.02409422232045068
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,5120,10240,0.023057777020666335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,12288,0.14641333950890437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,12288,0.017476444443066914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,65536,0.21715288692050508
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,5120,8192,0.020199111766285367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,5120,10240,0.022283555732833013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,10240,0.015509333875444202
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,5120,7168,0.01777688827779558
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,8192,0.1413431167602539
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,8192,0.01371022231049008
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,5120,8192,0.02105866703722212
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,6144,128,0.1366977824105157
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,5120,7168,0.019318222999572754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,5120,6144,0.016022221909628976
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,7168,0.14082133769989014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,6144,32,0.011404444773991903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,5120,6144,0.01872711049185859
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,6144,0.1473271052042643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,6144,0.011462222370836469
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,5120,0.010348444183667501
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,5120,0.1463342242770725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,5120,4096,0.01648177703221639
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,4096,0.14437599976857504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,5120,3584,0.009661333428488838
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,4096,0.008987555901209513
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,5120,3584,0.01672266589270698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,3584,0.14739999506208631
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,10240,0.1423964500427246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,3584,0.008656000097592672
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,5120,3072,0.015805333852767944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,3072,0.1461120049158732
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,5120,2560,0.00794488853878445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,7168,0.012831110921170978
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,3072,0.007988444632954067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,5120,5120,0.014282666974597506
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,5120,2048,0.007128000259399414
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,5120,2560,0.014582221706708273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,2560,0.14385688304901123
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,2560,0.007613333563009898
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,5120,2048,0.013756444056828817
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,2048,0.14048266410827637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,5120,4096,0.011100444528791638
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,2048,0.007302222152551015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,5120,1024,0.004969777746333016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,5120,1536,0.013106666505336761
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,1536,0.14220710595448813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,1536,0.006847111301289664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,5120,768,0.0042408886882993905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,1024,0.13933066527048746
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,1024,0.006238222122192383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,5120,1024,0.012498666842778524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,5120,512,0.003953777667548922
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,768,0.005951111101441913
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,768,0.1394586695565118
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,5120,768,0.012503999802801343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,5120,3072,0.008801778157552084
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,5120,256,0.0032604444358083936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,512,0.005883555445406172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,512,0.14057332939571804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,5120,512,0.011853333148691388
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,5120,128,0.0029582221888833572
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,5120,256,0.011397333608733283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,256,0.0058959999846087555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,5120,64,0.0028213332924577924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,5120,128,0.005599110904667113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,128,0.1323439942465888
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,5120,32,0.0028373334142896864
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,5120,256,0.29491111967298717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,5120,128,0.011507555842399597
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,5120,64,0.011402666568756104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,4096,65536,0.0889493359459771
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,5120,32,0.011400000088744693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,5120,1536,0.006219555520349079
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,5120,5120,0.017712000343534682
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,4096,16384,0.026670222481091816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,4096,12288,0.02126577827665541
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,4096,65536,0.07916355795330472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,4096,16384,0.028600000672870215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,16384,0.14025510682000056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,4096,10240,0.018606222338146634
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,4096,12288,0.022681777675946552
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,12288,0.01682133310370975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,12288,0.14176532957288954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,65536,0.18345067236158583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,4096,8192,0.016207999653286405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,4096,10240,0.021365332934591506
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,10240,0.14709866046905518
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,4096,7168,0.014749333262443542
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,8192,0.013088000317414602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,8192,0.14152533478207058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,4096,8192,0.019912888606389362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,4096,6144,0.013422222601042854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,4096,7168,0.018538667096032035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,7168,0.011928888658682505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,7168,0.14039644930097792
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,4096,6144,0.017903111047214933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,6144,0.1461377806133694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,6144,0.011130666567219628
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,4096,5120,0.016700444949997794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,5120,0.009420444567998251
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,4096,4096,0.010423999693658618
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,4096,4096,0.016061334146393668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,16384,0.02060622142420875
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,4096,0.14565689033932155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,4096,3584,0.009949333137936061
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,4096,0.008391111261314815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,4096,3584,0.015114666687117683
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,4096,3072,0.008996444443861643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,10240,0.015101333459218344
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,4096,3072,0.014112888111008538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,4096,2560,0.007987555530336168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,3072,0.1447395616107517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,4096,2560,0.013420444395807056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,2560,0.1447439988454183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,4096,2048,0.006710222197903528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,4096,5120,0.01218311074707243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,2560,0.007605333295133378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,4096,2048,0.013463111387358772
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,2048,0.007093333535724216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,4096,1536,0.005236444373925527
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,2048,0.13975732856326634
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,5120,0.14776621924506292
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,4096,1024,0.004210666649871402
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,1536,0.006607999818192587
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,4096,1536,0.013103110922707452
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,1536,0.1455964379840427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,65536,0.06770666440327962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,4096,1024,0.0121760004096561
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,1024,0.00628355559375551
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,1024,0.1396026611328125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,4096,512,0.0032275555034478507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,4096,768,0.012044444680213928
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,768,0.006271111054552927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,768,0.14738933245340982
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,3584,0.14145689540439182
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,4096,256,0.0031475555151700974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,4096,512,0.012080888781282636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,512,0.13981333043840197
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,4096,256,0.01146666705608368
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,3072,0.00794400026400884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,256,0.14554756217532688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,256,0.0058755555914507965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,4096,128,0.011398221883508893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,128,0.005927111125654644
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,4096,128,0.13759377267625597
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,4096,64,0.0028604444944196274
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,4096,32,0.002855111120475663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3584,65536,0.07571022378073798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,4096,32,0.011397333608733283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3584,65536,0.0776053335931566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,4096,768,0.0038559999730851916
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3584,16384,0.02401599950260586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,65536,0.06693689028422038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,65536,0.18404000335269502
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,3584,0.00811822215716044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3584,16384,0.026193777720133465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3584,12288,0.019483556350072224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,4096,512,0.005934222290913264
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,16384,0.020240000552601285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,16384,0.14871644973754883
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3584,10240,0.016809778081046212
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,4096,128,0.002867555452717675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3584,12288,0.022286223040686712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,12288,0.016806221670574613
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3584,8192,0.014917333920796713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3584,10240,0.020780444145202637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,10240,0.015014222926563688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3584,7168,0.013732444081041547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3584,8192,0.0195777780479855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,8192,0.14462310738033718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,8192,0.012454222473833295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,4096,64,0.011380444798204633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3584,6144,0.013078221844302284
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3584,7168,0.018246221873495314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,7168,0.011722666521867117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,7168,0.1442719962861803
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3584,5120,0.01202044470442666
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3584,6144,0.01738755570517646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,6144,0.01035822182893753
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,6144,0.14705688423580593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3584,4096,0.01055555542310079
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3584,5120,0.016792888442675274
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,5120,0.1443208853403727
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3584,3584,0.009396444592210982
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,4096,0.008579555484983657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,4096,0.14160533746083578
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3584,3584,0.014762666490342883
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,3584,0.00795733349190818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3584,3072,0.008589333130253686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,3584,0.1460568904876709
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3584,3072,0.01387288918097814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,3072,0.007979555262459649
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,3072,0.14517866240607366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3584,2560,0.007928888830873702
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,12288,0.14211911625332302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3584,2048,0.005869333528810077
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3584,2560,0.014154665999942355
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,10240,0.140428450372484
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3584,2048,0.013436444103717804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,2048,0.0069573331210348345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,2048,0.14499378204345703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3584,1536,0.01277955538696713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,1536,0.1458186705907186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3584,1024,0.004210666649871402
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,1536,0.006637333167923822
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,5120,0.010008888940016428
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3584,4096,0.015126221709781222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3584,1024,0.012411555482281579
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3584,768,0.003527111063400904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,1024,0.006263111200597551
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3584,768,0.012404444317022959
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3584,512,0.00332622230052948
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,768,0.006262222097979651
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3584,512,0.011776000261306763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,512,0.14071555932362875
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3584,256,0.003160888950030009
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,2560,0.007600000335110559
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,2560,0.1449822187423706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3584,128,0.0028151110228565005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3584,256,0.011865778101815118
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,256,0.1392995516459147
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,256,0.0057955553962124725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3584,1536,0.00490133340160052
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3584,128,0.011191999746693505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,128,0.0058577776783042485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3584,32,0.0028275555620590844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3584,64,0.011374221907721626
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3072,65536,0.06731466452280681
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3584,32,0.011364444262451597
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3072,65536,0.07597155703438653
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3072,16384,0.020968000094095867
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,65536,0.06593510839674208
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,1024,0.14153510994381374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,768,0.14613333013322619
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,16384,0.02015822297996945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3072,12288,0.01809955636660258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,65536,0.18292621771494547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,16384,0.14039822419484457
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3072,12288,0.022152000003390845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,12288,0.016511999898486666
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,12288,0.14733866850535074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3072,10240,0.015483554866578845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,3584,512,0.005917333480384614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3072,8192,0.01334755536582735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3072,10240,0.020403555697864957
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,10240,0.1462684472401937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3584,64,0.0025271111064487034
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3072,8192,0.0196079992585712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,8192,0.012123555772834353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,3584,128,0.1334364414215088
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3072,7168,0.01850666602452596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,7168,0.011577777564525604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3072,6144,0.011457777685589261
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,7168,0.14466488361358643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3072,6144,0.01716088917520311
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,6144,0.010335110955768162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,6144,0.14650132921006945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3072,5120,0.01110400011142095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3072,5120,0.01638488968213399
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3072,16384,0.025518221987618342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,5120,0.009472000102202097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3072,4096,0.010010666317409938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,5120,0.14619378248850504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3072,3584,0.009166222479608323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,4096,0.008390222158696916
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3072,4096,0.014134221606784396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,4096,0.13939200507269964
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3072,3584,0.01513777838812934
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,3584,0.00797155582242542
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,3584,0.14799466398027208
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,10240,0.01457244488928053
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3072,3072,0.013817778064144982
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3072,2560,0.007075555622577667
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,3072,0.007655111451943715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,3072,0.1423155599170261
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3072,7168,0.012759110993809171
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3072,2560,0.01349511080318027
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,2560,0.007287110719415877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,2560,0.14224355750613743
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3072,1536,0.004579555657174853
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3072,2048,0.01309866706530253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,2048,0.14516533745659724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,2048,0.006968888971540663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3072,1024,0.0038613333470291565
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3072,1536,0.012796444197495779
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,1536,0.1428177754084269
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,1536,0.0069546666410234236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3072,768,0.0038844446341196695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3072,1024,0.01205599970287747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,1024,0.006249777972698212
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,1024,0.14002044995625815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3072,512,0.003216888962520493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3072,768,0.01221333361334271
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,768,0.006077333456940121
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,768,0.14739378293355307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3072,256,0.0028328889360030494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3072,512,0.012082666986518435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,512,0.1447475618786282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,512,0.0058986664646201665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3072,3072,0.008243555823961893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3072,128,0.0028239999794297745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3072,256,0.011071110765139261
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,256,0.13976889186435276
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,256,0.005707555760939916
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3072,64,0.00282222218811512
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3072,128,0.011744889120260874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3072,32,0.0027919999427265594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,128,0.1328266726599799
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2560,65536,0.05901510847939385
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,3072,2048,0.005873777800136142
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3072,32,0.011016888750924004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,3072,8192,0.1403599977493286
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2560,16384,0.019561777512232464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2560,65536,0.07277688715193006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2560,16384,0.025348444779713947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,65536,0.06468266910976835
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,16384,0.019883554842736986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2560,12288,0.016161777906947665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,65536,0.18325066566467285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2560,12288,0.02162044412559933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,12288,0.01588088936275906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,12288,0.14147022035386828
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2560,10240,0.020272889071040683
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,10240,0.014087999860445658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2560,8192,0.012422222230169507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2560,8192,0.01865511139233907
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,8192,0.1426773336198595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2560,7168,0.011912888950771756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,8192,0.011355555719799466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,3072,128,0.005624889085690181
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2560,7168,0.01796000036928389
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,7168,0.010679110884666443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,7168,0.14308267169528538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2560,6144,0.011054221954610614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2560,5120,0.010496888723638324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,3072,64,0.011393778026103973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2560,6144,0.016996444927321542
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,6144,0.010039110978444418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,6144,0.1424942281511095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2560,5120,0.01538755496342977
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2560,4096,0.008953777452309927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,5120,0.14698044459025064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,5120,0.009342222577995723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2560,3584,0.008261333737108443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,16384,0.14003288745880127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,4096,0.008662222160233391
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2560,3584,0.013807111316257052
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2560,10240,0.014109333356221518
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,3584,0.14111733436584473
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2560,3072,0.013754666679435305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,3072,0.13920266098446316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2560,2560,0.006575999988449945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,3072,0.007840000092983246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2560,2560,0.013442666994200813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,2560,0.14090311527252197
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2560,2048,0.005573333137565189
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,2560,0.007303111255168915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2560,2048,0.013089777694808112
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,2048,0.14017066690656874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2560,1536,0.004555555681387584
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,2048,0.0069875551594628235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2560,4096,0.014104889498816596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,4096,0.1414817836549547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2560,1024,0.0041982221106688184
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,1536,0.14079022407531738
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,1536,0.006601777755551868
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2560,3072,0.007546666595670912
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2560,768,0.0036755556033717263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2560,1024,0.012056888805495368
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,1024,0.00628266649113761
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,10240,0.13966577582889134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2560,768,0.012057777908113269
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2560,512,0.0031946665710873077
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,768,0.14069422086079916
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2560,512,0.011746666497654386
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,512,0.0058782220714622075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,512,0.13938132921854654
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2560,256,0.0028648889727062653
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2560,256,0.011437333292431302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2560,128,0.0028275555620590844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2560,128,0.01108977778090371
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,128,0.13354221979777017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2560,1536,0.01275111072593265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2560,64,0.0028640000770489373
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,128,0.005776888794369168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,3584,0.008361777497662438
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2560,32,0.0028133332315418455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2560,64,0.011048888994587792
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2048,65536,0.04810133245256212
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,1024,0.14008355140686035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2560,32,0.011384000380833944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2048,16384,0.01575999955336253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,768,0.005964444329341252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2048,65536,0.0709644423590766
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,65536,0.06287644306818645
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2048,12288,0.013440888788965015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2048,16384,0.024517332514127094
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,16384,0.019318222999572754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,65536,0.18238488833109537
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,16384,0.14699822001987034
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2048,10240,0.01198933356338077
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2048,12288,0.021290666527218286
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,2560,256,0.005608888963858287
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,12288,0.14654133054945204
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,2560,256,0.13891022735171848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2048,10240,0.019766221443812054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,10240,0.1406559944152832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,10240,0.012999999854299756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2048,7168,0.012296888563368054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2048,8192,0.018191110756662156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,8192,0.011404444773991903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,8192,0.13988978332943386
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2048,6144,0.011724444727102915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2048,7168,0.01682400041156345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,7168,0.010951999988820819
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,7168,0.14152088430192736
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2048,6144,0.016502221425374348
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2048,5120,0.010755555497275459
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,6144,0.1464666657977634
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2048,4096,0.009858667022652095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2048,5120,0.015236443943447538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,5120,0.009361777868535783
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,5120,0.14016621642642552
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2048,3584,0.00833333366447025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2048,4096,0.014137778017255994
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,4096,0.008303999900817871
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,4096,0.14595911237928602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2048,3072,0.0069644442862934535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2048,3584,0.013776889277829064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,3584,0.007984889050324758
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,3584,0.14621867073906794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2048,2560,0.0058248887459437055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2048,3072,0.01348977784315745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,3072,0.007690666450394525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,3072,0.14277511172824436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2048,2048,0.00518488883972168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2048,2560,0.013447110851605734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,2560,0.007318221860461765
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,2560,0.14409866597917345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2048,2048,0.012823999755912356
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2048,1536,0.004575110971927643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,2048,0.13909599516126844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,12288,0.015533333023389181
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,2048,0.006948444578382704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2048,8192,0.012914666699038612
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2048,1536,0.012740444805887012
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,1536,0.14474933677249485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2048,768,0.0038373333712418876
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2048,1024,0.012099555797047086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,1024,0.14615021811591253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,1024,0.006239111224810283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2048,512,0.003165333428316646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2048,768,0.012091555529170565
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,768,0.1470924483405219
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,768,0.005952000204059813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2048,256,0.003160888950030009
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2048,512,0.011773332953453064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,512,0.005828444328573015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,512,0.1415111091401842
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2048,128,0.002846222163902389
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2048,256,0.011686222420798408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,256,0.14101155598958334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,256,0.005582222094138463
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2048,64,0.0028417776856157514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2048,128,0.011754666765530905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,128,0.005620444400442972
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,2048,128,0.13626844353146023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2048,64,0.010355555348926121
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1536,65536,0.03683022326893277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,2048,32,0.011074666347768573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1536,65536,0.06869866450627644
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1536,16384,0.013528000149461957
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,65536,0.06053511301676432
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2048,1024,0.0038782221575578055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,65536,0.18080711364746094
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,1536,0.006631111105283101
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1536,12288,0.011716444459226398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1536,16384,0.02405244443151686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,16384,0.018519111805491976
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,16384,0.14171111583709717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1536,12288,0.020837333467271592
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,12288,0.01399466726515028
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,12288,0.14485955238342285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1536,10240,0.01925155520439148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1536,8192,0.014722666806644864
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,2048,6144,0.00999111102686988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,10240,0.14047200149959987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1536,7168,0.013953778478834363
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1536,8192,0.017352888981501263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,8192,0.01110755569405026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,8192,0.14450132846832275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1536,6144,0.012607999973826937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1536,7168,0.01681422193845113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,7168,0.010423111418883005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,7168,0.14292800426483154
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,2048,32,0.0028275555620590844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1536,5120,0.011355555719799466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1536,6144,0.015815110670195687
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,6144,0.009858667022652095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,6144,0.14005333847469753
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1536,4096,0.009668444593747457
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,5120,0.009302222066455418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,5120,0.14201422532399496
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1536,3584,0.008979555633332994
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1536,4096,0.014093332820468478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,4096,0.00832444429397583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1536,3584,0.013800889253616333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,3584,0.1421120034323798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1536,3072,0.006238222122192383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,3584,0.007967111137178209
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1536,10240,0.01663555536005232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1536,3072,0.013609777722093793
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1536,2560,0.005580444302823808
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,3072,0.007656888829337225
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,3072,0.1433511045244005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1536,2048,0.005604444278611078
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,2560,0.007636444436179266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,10240,0.012763555679056378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1536,1536,0.004543999830881755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,2048,0.1462737719217936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,2048,0.0069395556218094295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1536,1536,0.012395555774370829
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,1536,0.0069395556218094295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1536,1024,0.012447111308574677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1536,5120,0.015106666419241162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,1024,0.006222222414281633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,1024,0.14547288417816162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,4096,0.14157511128319636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1536,768,0.012072000238630505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,768,0.14100978109571669
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1536,512,0.003504000190231535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,768,0.006103111224042044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1536,512,0.01202933324707879
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1536,2560,0.013370667066839008
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,2560,0.13993422190348306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1536,256,0.0029200000895394217
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,512,0.14548888471391466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,512,0.0058782220714622075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1536,256,0.011703111231327057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1536,128,0.0028488888508743713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,256,0.005578666511509154
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1536,1024,0.0038488888078265717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1536,128,0.011345778074529437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1536,64,0.0026311110705137253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,128,0.1317982276280721
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1536,32,0.0026400000270869997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1536,64,0.011352889239788055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1536,32,0.011016888750924004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1024,65536,0.027089777919981215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1536,768,0.0035457776652442086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1024,65536,0.06579822301864624
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,65536,0.05866577890184191
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1024,16384,0.013410666750537025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,65536,0.1809297800064087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1024,16384,0.023808000816239253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,16384,0.13924888769785562
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1024,12288,0.011012444065676795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1536,2048,0.01273333364062839
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1024,12288,0.020222221811612446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,12288,0.014668444792429606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,12288,0.14186399512820774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,256,0.1397955550087823
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,1536,1536,0.14490577909681532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1024,10240,0.018413333429230582
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1024,8192,0.009853333234786987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,10240,0.14178133010864258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,1536,128,0.00562755556570159
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1024,7168,0.008656000097592672
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,8192,0.011343110766675739
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,8192,0.14470133516523573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1024,7168,0.015899555550681222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,7168,0.010381333529949188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,7168,0.14232088459862605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1024,6144,0.015504888362354703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,6144,0.14317777421739367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1024,5120,0.007595555649863348
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,6144,0.009696000152164036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,16384,0.016840888394249808
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1024,5120,0.015096889601813423
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1024,4096,0.007287110719415877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,5120,0.14374489254421657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,5120,0.009301332963837517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1024,4096,0.014206222361988492
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1024,3584,0.007260444263617198
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,4096,0.00830844458606508
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,4096,0.14487822850545248
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,10240,0.012727999852763282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,3584,0.00795022232664956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,3584,0.14101955625745985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1024,8192,0.016712889075279236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1024,3072,0.013774221969975365
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1024,2560,0.00590844452381134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,3072,0.007631111476156447
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1024,2560,0.013226666384273104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1024,6144,0.007976888782448238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,2560,0.007473777565691207
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1024,2048,0.005580444302823808
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1024,2048,0.012539555629094442
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,2048,0.0069475554757648045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,2048,0.1420631143781874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1024,1536,0.012430222498046027
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,1536,0.14092532793680826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1024,1024,0.0041831110914548235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,1536,0.006618666566080517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1024,1024,0.012039110892348819
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,1024,0.1396791140238444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1024,768,0.003846222327815162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1024,3072,0.006292444550328785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1024,768,0.012083555261294046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,3072,0.14364177650875515
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1024,512,0.003488000068399641
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,768,0.14428888426886663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,768,0.006234666539563074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,2560,0.14083911312950984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1024,256,0.0031591111587153543
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1024,512,0.011740444435013665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,512,0.00590133335855272
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,512,0.14493155479431152
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1024,128,0.0029208889851967492
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1024,1536,0.0052560000783867305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1024,256,0.011359111302428775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,256,0.005618666609128316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,256,0.14444355169932047
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1024,64,0.002805333377586471
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1024,32,0.002874666617976295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,128,0.005622222191757626
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,1024,128,0.1393804417716132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,768,65536,0.02240177823437585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1024,64,0.0110871113008923
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1024,32,0.011014222270912595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,768,16384,0.015102222561836243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,768,65536,0.06273600127961901
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,65536,0.058442665470970996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,65536,0.1802471081415812
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,768,16384,0.022967111733224656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,1024,1024,0.006260444306664997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,16384,0.01681600014368693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1024,3584,0.013796444568369122
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,768,12288,0.01960444450378418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,12288,0.1395039955774943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,768,10240,0.009332444104883406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,768,10240,0.018163555198245578
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,10240,0.14000177383422852
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,768,8192,0.008757332960764566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,10240,0.012732444538010491
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,768,8192,0.016859556237856548
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,8192,0.13926045099894205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,1024,128,0.011615110768212212
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,768,7168,0.008287111090289222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,8192,0.011114666859308878
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,7168,0.010685332947307162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,768,7168,0.01641688909795549
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,7168,0.14080799950493708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,768,6144,0.007582222421964009
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,1024,10240,0.010302222437328762
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,768,5120,0.007247111035717859
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,768,6144,0.015499555402331881
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,768,12288,0.010105777945783403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,6144,0.14084355036417642
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,768,4096,0.0075733330514695905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,16384,0.14770132965511745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,768,5120,0.014820444915029736
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,5120,0.009311111436949836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,5120,0.14566044012705484
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,768,3584,0.007261333366235097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,768,4096,0.014035556051466199
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,4096,0.008330666356616551
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,4096,0.1462177832921346
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,768,3072,0.0069226668112807805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,768,3584,0.01385244478782018
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,3584,0.007978666987684038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,3584,0.14265422026316324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,768,2560,0.006887110984987683
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,768,3072,0.01310400002532535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,3072,0.14151022169325086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,3072,0.007738666401969061
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,768,2048,0.005199999858935674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,768,2560,0.013089777694808112
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,2560,0.007278222176763747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,2560,0.140864888827006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,768,1536,0.004559111264016894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,768,2048,0.013082666529549493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,2048,0.006978666616810693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,2048,0.14173155360751682
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,768,1024,0.0039048890272776284
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,768,1536,0.012425777812798819
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,1536,0.006683555742104848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,1536,0.14669511053297254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,768,768,0.003563555578390757
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,1024,0.006275555739800136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,1024,0.14177600542704263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,768,512,0.003212444484233856
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,768,768,0.011725333001878528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,768,0.006188444379303191
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,768,0.14612000518374973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,6144,0.009979555176364051
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,768,512,0.011755555868148804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,512,0.0058755555914507965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,512,0.14587999714745417
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,768,128,0.0028595555987623003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,768,256,0.011697777443461947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,256,0.005570666657553778
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,256,0.14107555813259548
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,12288,0.01407999959256914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,768,64,0.0028417776856157514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,768,128,0.011023999916182624
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,768,128,0.005601777798599667
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,768,128,0.1318382157219781
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,512,65536,0.0294773346847958
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,768,64,0.011182222101423474
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,768,32,0.01035733355416192
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,512,16384,0.010392000277837118
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,512,65536,0.06184089183807373
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,65536,0.05607999695671929
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,65536,0.17924711439344618
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,512,12288,0.008948444492287105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,512,16384,0.02199733257293701
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,16384,0.13923822508917913
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,16384,0.01685066686736213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,512,10240,0.007979555262459649
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,512,12288,0.0192586663696501
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,768,1024,0.01221155540810691
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,12288,0.014090667168299357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,512,8192,0.008583111067612966
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,512,10240,0.018241778016090393
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,10240,0.012727110750145383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,10240,0.14031733406914607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,512,7168,0.007663999994595845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,512,8192,0.01682044400109185
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,768,256,0.0028897778441508612
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,8192,0.011371555427710215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,512,6144,0.007297777467303806
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,512,7168,0.015832000308566622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,7168,0.13890311453077528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,512,6144,0.015462223026487561
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,512,5120,0.007254222200976477
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,6144,0.14040444956885442
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,512,5120,0.014780445231331719
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,5120,0.1443493366241455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,512,4096,0.007259555160999298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,5120,0.009341333475377824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,512,4096,0.013934221532609729
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,4096,0.14221955670250788
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,512,3584,0.007301333049933116
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,4096,0.008272888759771982
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,12288,0.14573332998487684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,512,3584,0.013791999883121915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,512,3072,0.0069395556218094295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,3584,0.007975111405054728
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,3584,0.14559022585550943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,8192,0.14329422844780818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,512,3072,0.013438222308953604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,3072,0.00776711106300354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,3072,0.14730666743384466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,7168,0.010440000229411654
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,512,2048,0.00627377794848548
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,512,2560,0.01312800000111262
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,2560,0.007283555964628856
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,2560,0.1449164417054918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,768,32,0.002528888897763358
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,512,1536,0.004232888834344016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,512,2048,0.013080889152155982
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,2048,0.0069120000633928515
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,2048,0.14173866642846003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,512,1024,0.003911111089918348
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,512,1536,0.012400888734393649
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,1536,0.006628444625271692
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,512,768,0.003539555602603488
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,512,1024,0.012379555238617791
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,1024,0.006252444452709622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,1024,0.146250671810574
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,512,512,0.0032400000426504347
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,512,768,0.012129777835475074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,768,0.006238222122192383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,768,0.14167911476559109
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,512,256,0.0031475555151700974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,512,512,0.011395555403497485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,512,0.005896889087226655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,512,0.14594577418433294
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,512,256,0.011677333050303988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,256,0.005602666487296422
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,256,0.14536355601416692
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,512,128,0.011352000137170156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,128,0.005599110904667113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,512,64,0.002790222151411904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,512,2560,0.006640000061856375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,512,32,0.0028702221396896574
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,256,65536,0.015285332997639975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,512,64,0.010035555395815108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,512,32,0.011353777514563667
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,512,6144,0.00998488896422916
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,256,65536,0.06064089139302572
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,256,16384,0.009000889129108852
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,65536,0.17968089050716826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,65536,0.05098311106363932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,256,12288,0.008325333396593729
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,256,16384,0.02195466723706987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,16384,0.1410142183303833
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,16384,0.016701334052615695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,256,10240,0.007646222081449296
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,256,12288,0.019557333654827543
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,12288,0.013996443814701505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,12288,0.14159644974602595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,256,8192,0.006963555183675554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,256,10240,0.017694221602545846
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,10240,0.012747555143303342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,10240,0.14411732885572645
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,256,7168,0.0069751110341813825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,256,8192,0.01681777834892273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,512,128,0.002834666727317704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,8192,0.14782044622633192
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,256,6144,0.006578666468461354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,256,7168,0.016162667009565566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,7168,0.14134577910105386
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,7168,0.010527110762066312
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,256,5120,0.006285333385070165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,256,6144,0.015142222245534262
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,128,0.1327022181616889
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,6144,0.00980533328321245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,256,4096,0.005929777605666055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,256,5120,0.014808000789748298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,5120,0.009311111436949836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,5120,0.14745599693722194
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,256,3584,0.006883555402358373
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,256,4096,0.014087110757827759
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,4096,0.008286221987671321
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,4096,0.14442488882276747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,256,3584,0.013444444371594323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,256,3072,0.005982222242487802
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,512,1536,0.14597066243489584
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,3584,0.14163111315833196
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,256,2560,0.006295111030340195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,256,3072,0.013087111214796701
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,3072,0.007647111184067196
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,256,2048,0.006052444378534953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,256,2560,0.013167999684810638
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,2560,0.007274666594134436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,2560,0.13928622669643825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,256,2048,0.01279022213485506
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,2048,0.1402319934633043
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,256,1536,0.004521777646409141
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,2048,0.0069724445541699724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,8192,0.011053333679835001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,256,1024,0.0036559998989105225
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,256,1536,0.012703999876976013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,1536,0.006674666785531574
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,1536,0.14497600661383733
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,256,1024,0.01238044434123569
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,256,768,0.0034924443397257063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,6144,0.14362933900621203
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,1024,0.006232000059551663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,1024,0.14065689510769314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,256,512,0.0031866667171319327
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,256,768,0.011738667057620155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,768,0.14149955908457437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,768,0.005945777727497949
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,256,256,0.0028515555378463534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,256,512,0.012055111428101858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,512,0.13999911149342856
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,256,128,0.002858666703104973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,256,256,0.011716444459226398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,256,0.13797067271338567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,3584,0.007956444389290279
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,256,64,0.002850666642189026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,256,128,0.011111111276679568
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,256,32,0.002530666689078013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,128,0.005597333527273602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,3072,0.1448684400982327
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,256,64,0.011086222198274402
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,256,32,0.011530666715568967
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,128,65536,0.058020446035597056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,128,16384,0.008036444584528605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,65536,0.05005688799752129
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,128,16384,0.022230222821235657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,16384,0.14101066854265001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,128,12288,0.007598222129874759
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,16384,0.01683288812637329
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,128,12288,0.019208000765906442
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,12288,0.1445280048582289
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,512,0.005904888941182031
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,12288,0.013873777455753751
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,256,256,0.005592888842026393
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,128,10240,0.018073777357737224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,10240,0.14148888323042128
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,128,8192,0.007000889215204451
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,256,128,0.13389155599806044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,10240,0.012727110750145383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,128,65536,0.014471999473041959
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,128,8192,0.016580444243219163
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,8192,0.011071110765139261
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,8192,0.14600088861253527
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,128,6144,0.006568888823191325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,128,7168,0.016142222616407607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,7168,0.010678222609890832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,7168,0.13982578118642172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,65536,0.17926222748226592
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,128,5120,0.006238222122192383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,128,6144,0.015318221516079374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,6144,0.009922666682137383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,6144,0.1402684450149536
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,128,4096,0.005928888916969299
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,128,5120,0.01479111115137736
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,5120,0.00925333301226298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,128,3584,0.006226666685607698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,128,4096,0.014117333624098035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,4096,0.008302222523424361
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,4096,0.14215555455949572
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,128,3072,0.0059368887709246735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,128,3584,0.013758222262064615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,3584,0.00793511089351442
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,3584,0.1403599977493286
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,128,2560,0.00655644428398874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,128,3072,0.013726222018400827
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,128,10240,0.007076444725195567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,3072,0.007662222617202335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,3072,0.14547288417816162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,128,2048,0.006113777971929974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,128,2560,0.013109332985348172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,2560,0.00741155528359943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,2560,0.1416320006052653
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,128,1536,0.004247111164861255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,128,2048,0.012714666624863943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,2048,0.006990222467316522
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,2048,0.1408862272898356
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,128,7168,0.006650666809744305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,128,1536,0.01274311128589842
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,1536,0.006628444625271692
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,128,768,0.003474666840500302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,128,1024,0.012040889097584618
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,1024,0.006224000205596288
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,128,768,0.012055111428101858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,768,0.14177244239383274
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,128,512,0.003149333306484752
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,128,512,0.011711111499203576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,5120,0.1410542196697659
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,512,0.14428088400099012
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,128,256,0.0028151110228565005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,512,0.005912000106440649
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,128,256,0.011586666935020022
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,128,128,0.0025173332542181015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,256,0.1406373315387302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,256,0.00556355549229516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,128,128,0.010689777632554373
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,128,64,0.002551111082235972
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,128,0.005616000129116907
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,128,0.1378248929977417
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,128,32,0.002647111192345619
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,64,65536,0.010660444696744284
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,128,64,0.010017777482668558
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,128,1024,0.003810666501522064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,128,32,0.011020444333553314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,64,12288,0.006285333385070165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,1536,0.14571821689605713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,64,65536,0.05724355247285631
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,64,10240,0.006259555617968242
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,64,16384,0.021925333473417494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,64,8192,0.006641777853171031
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,64,12288,0.0192284451590644
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,64,7168,0.006594666590293248
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,64,6144,0.0063075555695427795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,64,10240,0.017892445127169292
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,64,8192,0.016516443755891588
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,64,5120,0.005913777897755305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,64,7168,0.015856888559129503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,64,4096,0.005613333235184352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,64,3584,0.006233777850866318
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,64,5120,0.014819555812411837
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,64,4096,0.013803555733627744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,64,3072,0.0058746664888328975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,64,2560,0.006288888967699475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,64,3072,0.013411555853154926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,64,2560,0.013105777402718862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,64,2048,0.01275466630856196
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,64,1536,0.004516444272465176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,64,1536,0.012444444828563266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,64,1024,0.0038133333954546186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,32,128,1024,0.13988533284929064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,64,1024,0.012059555285506778
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,64,768,0.0035084444615576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,64,512,0.003129777808984121
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,32,128,768,0.005992888990375731
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,64,256,0.002805333377586471
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,64,768,0.012030222349696688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,64,128,0.002528888897763358
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,64,512,0.01202755586968528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,64,6144,0.015487111277050443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,64,64,0.0024702221983008915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,64,256,0.011705777711338468
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,64,128,0.011402666568756104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,64,3584,0.013744889034165276
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,64,64,0.011400000088744693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,32,65536,0.00928177767329746
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,64,32,0.011355555719799466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,32,16384,0.00692622239391009
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,32,12288,0.006227555374304454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,64,2048,0.005902222047249476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,32,10240,0.006226666685607698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,32,65536,0.05681955814361572
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,32,16384,0.02199111051029629
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,32,8192,0.005612444546487596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,32,12288,0.019573332534896005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,32,7168,0.006230222268237009
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,32,6144,0.0062977779242727495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,32,10240,0.018239110708236694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,32,8192,0.01664799948533376
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,32,7168,0.016122667325867545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,32,6144,0.015475556254386902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,32,5120,0.014783999986118741
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,32,3072,0.00590133335855272
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,32,4096,0.014096000128322177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,64,16384,0.006996444529957241
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,32,2560,0.006256000035338932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,32,3584,0.013781333135233985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,32,2048,0.005889777921968036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,32,1536,0.004503111044565837
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,32,2560,0.01274400038851632
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,32,1024,0.003559999995761447
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,32,2048,0.012811555630630918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,32,1536,0.01240711079703437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,32,1024,0.011718221836619906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,32,512,0.0031946665710873077
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,32,768,0.012039110892348819
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,32,256,0.0028080000645584534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,64,32,0.002520888836847411
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,32,128,0.0027964444210131965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,32,512,0.01110577748881446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,32,64,0.002516444358560774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,32,128,0.011424000064531961
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,32,32,0.0028159999185138275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,32,64,0.010067555639478896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,32,5120,0.005897777775923411
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,32,4096,0.005616000129116907
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,32,3584,0.005936000082227919
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,65536,16384,0.32378845744662815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,65536,16384,0.1712311108907064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,32,3072,0.013408000270525614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,65536,12288,0.2277688980102539
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,32,32,768,0.003571555432346132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,65536,12288,0.12985866599612766
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,16384,0.1569813357459174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,32,256,0.01146044416560067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,65536,10240,0.18927466869354248
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,65536,10240,0.11202311515808105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,65536,8192,0.1550275617175632
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,32,32,32,0.011363555159833698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,10240,0.09666222333908081
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,65536,8192,0.09679288996590509
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,65536,7168,0.13410666253831652
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,12288,0.24111554357740614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,8192,0.07215378019544813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,16384,0.29925333129035103
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,65536,7168,0.09697955846786499
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,65536,6144,0.11549066172705756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,10240,0.2071902222103543
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,7168,0.06392355759938557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,8192,0.176073776351081
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,65536,6144,0.08516266610887314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,12288,0.11190577348073323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,65536,5120,0.09728799925910102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,65536,4096,0.0807795524597168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,6144,0.056114666991763644
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,65536,5120,0.07669422361585829
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,5120,0.047747555706236094
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,7168,0.2043759955300225
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,65536,3584,0.06945600112279256
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,4096,0.039155556095971
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,6144,0.16797867086198595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,5120,0.15206489298078749
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,65536,3584,0.05730577972200182
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,65536,3072,0.06035911374621921
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,3584,0.03495288888613383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,65536,2560,0.051036443975236684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,3072,0.031705776850382485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,3584,0.14701333310869005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,65536,3072,0.05065600077311198
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,65536,2560,0.043996443351109825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,65536,2048,0.042043556769688926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,3072,0.14336800575256348
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,2560,0.027394667267799377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,65536,1536,0.032615999380747475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,65536,2048,0.037096000379986234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,2560,0.14416889349619547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,2048,0.023339556323157415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,65536,1536,0.03492355677816603
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,2048,0.14170400301615396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,1536,0.02216177847650316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,65536,1024,0.02340000040001339
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,65536,768,0.018894222047593858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,1024,0.1454275581571791
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,1024,0.0151591118839052
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,65536,4096,0.06492533286412557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,65536,512,0.013339555925793119
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,65536,768,0.020604444874657523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,768,0.14627110958099365
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,65536,256,0.00830577810605367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,65536,512,0.01891288823551602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,768,0.01349511080318027
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,512,0.14124888843960232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,4096,0.15907111432817247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,65536,256,0.015784889459609985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,256,0.010638222098350525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,256,0.14436977439456516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,65536,128,0.014440000057220459
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,65536,64,0.00589155571328269
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,128,0.1359297831853231
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,65536,32,0.0059279998143513995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,65536,64,0.015430221954981485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,65536,32,0.016200888488027785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,65536,1024,0.02316444449954563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,16384,65536,0.32313866085476345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,65536,1536,0.14264978302849665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,16384,16384,0.09148799710803562
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,16384,16384,0.05677511294682821
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,512,0.012757333616415659
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,65536,0.1552613311343723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,65536,128,0.006575999988449945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,16384,0.044870220952563815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,16384,0.14455466800265843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,16384,12288,0.04509688748253716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,65536,128,0.010580444501505958
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,12288,0.0351182222366333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,16384,10240,0.05245688888761732
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,16384,10240,0.04041066765785217
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,16384,8192,0.04487377736303541
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,12288,0.1428906652662489
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,10240,0.033095998896492854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,16384,7168,0.037971556186676025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,16384,8192,0.0355697770913442
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,10240,0.1456462277306451
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,8192,0.024724443753560383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,16384,6144,0.03363022208213806
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,8192,0.1447075605392456
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,16384,65536,0.19115822845035127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,16384,7168,0.03901422354910109
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,7168,0.02789688772625393
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,16384,5120,0.028576887316173975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,7168,0.1426408953136868
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,16384,6144,0.03649155630005731
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,16384,5120,0.031523555517196655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,6144,0.144214219517178
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,16384,4096,0.02427111069361369
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,5120,0.018163555198245578
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,5120,0.14337333043416342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,16384,3584,0.02089866664674547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,16384,4096,0.026492445005310908
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,4096,0.015465777781274585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,4096,0.1454355584250556
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,16384,3072,0.01869155632125007
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,16384,3584,0.02443733314673106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,3584,0.014469333820872836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,16384,3072,0.022433777650197346
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,16384,2560,0.01587288909488254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,3072,0.14185421996646458
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,16384,2560,0.020620443754725985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,2560,0.14132355319129095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,16384,2048,0.013063111239009433
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,2560,0.012359110845459832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,16384,2048,0.019208888212839764
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,65536,0.22743021117316353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,2048,0.1446053319507175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,16384,1536,0.01074577785200543
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,2048,0.011056000159846412
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,16384,1536,0.01792711185084449
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,1536,0.009296889106432596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,16384,1024,0.008255999949243333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,1536,0.14861244625515407
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,16384,1024,0.014928888943460254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,1024,0.14358577463361952
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,16384,768,0.006996444529957241
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,3584,0.1456915537516276
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,16384,768,0.014833778142929077
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,768,0.007304888632562425
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,3072,0.013231111069520315
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,16384,512,0.013403555585278405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,512,0.1384399996863471
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,16384,256,0.0032106666929192017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,512,0.007244444555706448
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,16384,256,0.012219555675983429
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,6144,0.02015911042690277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,256,0.14454844262864855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,16384,128,0.0031822222388452957
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,256,0.006641777853171031
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,128,0.00673333348499404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,16384,128,0.011679111255539788
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,16384,64,0.002805333377586471
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,16384,32,0.0028106667515304354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,16384,64,0.012079111403889127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,16384,1024,0.008040889269775814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,16384,32,0.01311288856797748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,768,0.14058844248453775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,12288,65536,0.25552532407972545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,12288,65536,0.16227732764350042
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,12288,16384,0.06760355499055651
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,12288,16384,0.05011644297175937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,65536,0.1482026709450616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,16384,0.03967644439803229
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,16384,0.14489955372280544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,16384,128,0.13130844963921443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,16384,12288,0.07179822524388631
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,12288,10240,0.04495911134613884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,12288,0.03094489044613308
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,12288,10240,0.03625511129697164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,10240,0.028167999453014795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,12288,8192,0.03762755460209317
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,16384,512,0.005587555468082428
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,10240,0.14271999730004206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,12288,7168,0.03310577736960517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,8192,0.022649778260125056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,8192,0.1388275490866767
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,12288,7168,0.035360889302359685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,12288,6144,0.029478222131729126
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,7168,0.020469332734743755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,7168,0.14553956190745035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,12288,12288,0.05270488725768196
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,12288,6144,0.03502488798565335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,6144,0.018560000591807894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,6144,0.14113777213626438
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,12288,12288,0.040234665075937905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,12288,5120,0.028984887732399836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,12288,4096,0.022582223018010456
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,5120,0.01683111157682207
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,5120,0.1461226675245497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,12288,0.1437022156185574
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,12288,3584,0.019627556204795837
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,12288,4096,0.025565332836574976
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,4096,0.01441422270403968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,4096,0.14536266856723362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,12288,3072,0.017714665995703805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,65536,0.18957244025336373
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,3584,0.1450844473308987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,3584,0.013379555609491138
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,12288,3584,0.02290577855375078
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,12288,2560,0.01240888900227017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,3072,0.13956799772050646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,12288,8192,0.0324524442354838
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,3072,0.012409778104888068
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,12288,2048,0.010375111467308467
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,12288,2560,0.02034311162100898
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,2560,0.011391999820868174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,2560,0.14513688617282444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,12288,1536,0.009304000271691216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,12288,2048,0.017878222796652053
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,2048,0.010039110978444418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,2048,0.1413653294245402
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,12288,1024,0.00758755538198683
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,12288,1536,0.016220443778567843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,1536,0.14413421683841282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,1536,0.008644444247086843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,12288,768,0.006270222365856171
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,12288,1024,0.0144213338692983
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,1024,0.007644444704055786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,1024,0.14231022198994955
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,12288,512,0.004922666483455234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,12288,5120,0.025587555434968736
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,768,0.007132444116804335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,768,0.1460391150580512
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,12288,256,0.0035439998739295532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,512,0.0069431112044387394
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,512,0.14313777287801108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,12288,128,0.0031591111587153543
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,12288,256,0.012207110722859701
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,256,0.140374223391215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,12288,128,0.011695110963450538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,12288,128,0.13727022541893855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,128,0.006846222198671765
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,12288,32,0.00314666661951277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,12288,64,0.012095111111799875
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,12288,3072,0.020987555384635925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,12288,32,0.012727110750145383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,10240,65536,0.21764443980322945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,10240,65536,0.15325600571102566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,10240,16384,0.06022310919231839
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,12288,768,0.01311466677321328
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,12288,512,0.013122667041089801
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,10240,16384,0.04814222123887804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,65536,0.13967822657691106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,10240,12288,0.0473271111647288
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,12288,256,0.006575999988449945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,16384,0.041004445817735456
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,10240,12288,0.03930222325854831
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,16384,0.14082133769989014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,10240,10240,0.03882488939497206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,12288,64,0.002855111120475663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,12288,0.029504001140594482
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,10240,8192,0.0328053335348765
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,12288,0.1465537812974718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,10240,10240,0.03533600105179681
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,65536,0.17896710501776802
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,10240,7168,0.02951999836497837
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,10240,0.141576886177063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,10240,8192,0.03148977624045478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,8192,0.021493333909246657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,10240,6144,0.02638133366902669
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,8192,0.14184089501698813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,7168,0.01996977792845832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,10240,7168,0.030557334423065186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,7168,0.14571911758846706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,10240,5120,0.023009777069091797
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,6144,0.01793688866827223
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,10240,6144,0.026291555828518335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,6144,0.14234933588239881
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,10240,4096,0.019712888532214694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,10240,5120,0.02459377712673611
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,5120,0.016157334049542744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,10240,3584,0.01755288905567593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,5120,0.14564889007144505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,10240,4096,0.022248889009157818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,4096,0.14096444182925755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,10240,3072,0.015771556231710646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,10240,3584,0.02125511070092519
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,3584,0.14225155777401394
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,10240,2560,0.012456888953844706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,3072,0.012053333222866058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,3072,0.14631111092037624
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,10240,2560,0.0188417782386144
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,10240,2048,0.011023110813564725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,2560,0.011041777829329172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,2560,0.13812888993157282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,2048,0.009691555466916826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,2048,0.14739999506208631
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,10240,0.026368000441127356
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,10240,1536,0.008657777474986183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,1536,0.008443555898136562
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,1536,0.1412737766901652
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,10240,1024,0.0075697774688402815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,10240,1024,0.013817778064144982
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,1024,0.1393315527174208
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,10240,768,0.005886222339338726
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,1024,0.0074133334888352295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,4096,0.014022222823566861
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,10240,768,0.01310488912794325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,10240,512,0.004534222185611725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,768,0.1396426624721951
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,3584,0.013215111361609565
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,10240,256,0.00388355553150177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,10240,512,0.012899555265903473
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,512,0.006758222149478064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,512,0.14361333847045898
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,10240,256,0.012124444047609964
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,10240,128,0.00318755561278926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,256,0.0063822223908371395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,10240,128,0.011598221957683563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,128,0.13604622417026097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,10240,64,0.0028106667515304354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,128,0.006629333313968446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,10240,32,0.0031840000301599503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,10240,1536,0.015787555111779105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,10240,64,0.011689778003427716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,10240,32,0.012470222181744046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,8192,65536,0.17466399404737684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,8192,65536,0.10197244750128852
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,8192,16384,0.04606399933497111
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,10240,768,0.0069546666410234236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,10240,3072,0.01975911193423801
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,65536,0.08104533619350858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,8192,12288,0.03276355399025811
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,10240,256,0.13895821571350098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,16384,0.024217777782016333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,16384,0.14295910464392766
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,8192,10240,0.027795556518766616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,10240,2048,0.017492444978819955
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,12288,0.01919911139541202
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,65536,0.17650311523013643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,12288,0.13959644900427923
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,8192,10240,0.02774488925933838
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,8192,8192,0.023805333508385554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,10240,0.01812622282240126
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,10240,0.14245333936479357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,8192,7168,0.020661332541041903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,8192,8192,0.023855111665195886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,8192,0.01479466590616438
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,8192,6144,0.018208889497650992
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,7168,0.014136888914638095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,7168,0.14209155241648355
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,8192,6144,0.023268444670571223
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,6144,0.013064000341627332
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,6144,0.13893155256907144
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,8192,16384,0.036650667587916054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,8192,5120,0.0207360006040997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,5120,0.13775732782151964
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,5120,0.011894221935007306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,8192,4096,0.019931554794311523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,8192,12288,0.03004177742534214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,4096,0.14465510845184326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,8192,3584,0.01205244412024816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,4096,0.011038222246699862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,3584,0.009966221948464712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,8192,3584,0.018231999542978074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,8192,3072,0.010712888505723743
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,8192,0.14234756098853216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,8192,3072,0.01719999975628323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,8192,7168,0.025352888637118872
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,8192,2560,0.009711110757456886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,3072,0.14152355988820395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,3072,0.008950221869680617
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,8192,2048,0.008267555799749162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,2560,0.008613333106040955
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,2560,0.14435911178588867
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,8192,2048,0.015525332755512662
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,2048,0.007713778151406183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,8192,1536,0.007623111208279927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,8192,4096,0.013432888521088494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,8192,1536,0.013777777552604675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,1536,0.1414195564058092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,8192,1024,0.0058702222175068325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,1536,0.0069751110341813825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,8192,1024,0.013094222380055321
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,1024,0.006593777901596493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,8192,768,0.004564444637960858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,8192,768,0.0123831108212471
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,3584,0.14566400316026476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,768,0.13867644468943277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,8192,512,0.003555555724435382
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,8192,512,0.011759999725553723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,512,0.006063999815119638
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,512,0.13919911119672987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,8192,256,0.01148444414138794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,8192,2560,0.01684977776474423
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,256,0.13897955417633057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,8192,5120,0.015822221835454304
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,256,0.005893333504597346
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,8192,128,0.011391999820868174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,2048,0.14021510548061794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,128,0.13675822152031794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,128,0.00590133335855272
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,8192,32,0.002883555574549569
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,8192,64,0.011376000112957425
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,8192,32,0.012053333222866058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,7168,65536,0.15017599529690212
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,8192,1024,0.14192977216508654
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,7168,65536,0.09553333123524983
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,7168,16384,0.042189333173963756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,8192,768,0.0062773335311147906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,7168,16384,0.03437866767247518
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,7168,12288,0.03338933322164748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,16384,0.023137778043746948
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,8192,256,0.0032008888406885993
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,16384,0.14317688677046034
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,7168,12288,0.028431998358832464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,7168,10240,0.028538667493396338
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,12288,0.1423555612564087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,65536,0.1752337747149997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,7168,10240,0.026087111896938745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,8192,64,0.0028044444819291434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,10240,0.1424364381366306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,10240,0.016509332590632968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,7168,8192,0.023637332850032385
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,8192,0.014442667365074158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,7168,7168,0.019183999962276883
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,8192,0.14184533225165472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,7168,7168,0.024155555499924555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,7168,0.013384888569513956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,7168,0.1395324468612671
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,7168,6144,0.016692444682121277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,65536,0.0751697752210829
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,6144,0.012381333443853589
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,7168,5120,0.01443822185198466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,7168,5120,0.020227554771635268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,5120,0.1460097763273451
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,7168,4096,0.012251555919647217
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,5120,0.011400889191362592
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,12288,0.018394667241308425
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,7168,4096,0.018903111418088276
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,4096,0.009492444495360056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,7168,3584,0.011071999867757162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,4096,0.14101866881052652
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,7168,8192,0.0221377776728736
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,7168,3584,0.018153778380817838
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,3584,0.009013333254390294
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,7168,3072,0.0099973330895106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,3584,0.14539999432033962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,7168,2560,0.008618666893906063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,7168,3072,0.01707466608948178
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,3072,0.1422035561667548
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,3072,0.008730666504965888
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,7168,2048,0.007695111135641734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,7168,2560,0.01614488826857673
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,2560,0.007954667011896769
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,2560,0.14406666490766737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,7168,1536,0.006903999795516332
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,7168,2048,0.014450665977266101
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,7168,6144,0.023294222023752
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,2048,0.007315555380450354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,6144,0.13974044058057997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,2048,0.14508177174462214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,7168,1024,0.005555555638339784
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,7168,1536,0.013756444056828817
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,1536,0.0068986668354935115
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,1536,0.1400737762451172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,7168,768,0.0041786668201287585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,7168,1024,0.012718222207493253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,1024,0.006441777778996362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,1024,0.1449120044708252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,7168,512,0.0037493333220481873
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,768,0.006235555642180973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,7168,768,0.012834666503800286
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,8192,128,0.0031315556002987754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,768,0.14089332686530218
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,7168,256,0.0031679999083280563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,7168,128,0.0028648889727062653
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,512,0.005943111247486538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,512,0.14457600646548802
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,7168,256,0.012064889073371887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,7168,64,0.0028080000645584534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,256,0.005879999862776862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,256,0.14016621642642552
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,7168,32,0.002847111059559716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,7168,128,0.01072088877360026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,7168,128,0.005881777654091517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,7168,128,0.1366542180379232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,7168,64,0.011397333608733283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,6144,16384,0.036017778846952654
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,6144,65536,0.1227911048465305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,6144,65536,0.08758399883906047
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,6144,12288,0.02904799911710951
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,6144,16384,0.03223199976815118
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,65536,0.07208177778455946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,16384,0.02195466723706987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,16384,0.1447653373082479
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,6144,10240,0.024923556380801733
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,6144,12288,0.02604622145493825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,12288,0.14369600349002415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,6144,10240,0.023034666975339253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,6144,8192,0.02156266735659705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,10240,0.01622488929165734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,10240,0.1406924459669325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,65536,0.1749324401219686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,6144,7168,0.019191111127535503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,6144,8192,0.02193333374129401
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,8192,0.013933334085676404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,8192,0.1420160002178616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,6144,6144,0.01720711092154185
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,7168,0.013073777986897362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,6144,7168,0.022309333086013794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,6144,5120,0.015200000670221118
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,7168,512,0.012102222277058495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,6144,6144,0.02233422299226125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,6144,0.14566755294799805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,6144,5120,0.01957244508796268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,5120,0.14207555188073054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,5120,0.011047110789351992
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,6144,4096,0.011698666546079846
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,7168,32,0.01202311118443807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,6144,3584,0.010008888940016428
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,6144,4096,0.01848088867134518
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,4096,0.009105777574910058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,4096,0.14601422680748835
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,6144,3072,0.009663999908500248
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,6144,3584,0.017582222819328308
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,3584,0.008647999829716152
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,3584,0.14222310649024114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,6144,2560,0.007985778152942657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,12288,0.01793066660563151
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,6144,3072,0.016740444633695815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,3072,0.008007110820876228
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,3072,0.13908178276485866
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,6144,2048,0.007284444239404466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,6144,2560,0.015816888875431485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,2560,0.0076142218377855085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,2560,0.13986666997273764
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,6144,1536,0.006643555644485686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,6144,2048,0.014120888378885059
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,2048,0.007283555964628856
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,2048,0.1476551161872016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,6144,1024,0.005201777650250329
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,6144,1536,0.01372977760103014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,1536,0.006903999795516332
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,1536,0.1454577843348185
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,6144,768,0.0042168887125121225
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,7168,0.141503996319241
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,6144,1024,0.01239377756913503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,1024,0.006568000134494569
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,6144,512,0.0035048888789282907
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,6144,768,0.012461333639091916
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,6144,0.012025777664449481
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,768,0.006259555617968242
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,768,0.1400302251180013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,6144,256,0.0031591111587153543
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,6144,512,0.012044444680213928
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,512,0.14755822552575007
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,512,0.005932444499598608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,6144,256,0.011723555624485016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,256,0.005616888817813661
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,256,0.1401217778523763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,6144,64,0.002854222224818336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,6144,32,0.002487999904486868
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,6144,128,0.010377777947319878
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,6144,128,0.005897777775923411
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,6144,64,0.011058666639857821
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,6144,32,0.011754666765530905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,5120,65536,0.07990044355392456
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,5120,16384,0.03223111232121786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,65536,0.06978578037685819
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,5120,16384,0.030349334081013996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,16384,0.14606577820248073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,16384,0.020996444755130347
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,65536,0.17527733908759224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,5120,12288,0.026318222284317017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,1024,0.1394586695565118
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,5120,10240,0.02235200007756551
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,5120,12288,0.024531554844644334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,12288,0.1400053368674384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,6144,128,0.0028782222006056043
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,5120,8192,0.019527110788557265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,5120,10240,0.022109333011839125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,10240,0.14070133368174234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,10240,0.015551111764378019
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,6144,128,0.13686488734351263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,5120,7168,0.017529777354664274
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,5120,8192,0.02084000077512529
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,8192,0.14095466666751438
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,8192,0.013520888984203339
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,5120,6144,0.015660444895426433
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,5120,7168,0.02163022259871165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,7168,0.012734221915404001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,7168,0.1414382192823622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,5120,5120,0.013817778064144982
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,5120,6144,0.020954666866196528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,6144,0.14157777362399632
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,6144,0.01168444421556261
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,5120,4096,0.010651555326249866
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,5120,5120,0.01920000049802992
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,5120,0.14109867148929173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,5120,0.010336000058386061
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,5120,3584,0.009909333454238044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,5120,4096,0.01790488925245073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,4096,0.14083733823564318
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,5120,3072,0.009083555804358589
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,3584,0.00830666638082928
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,3584,0.1400266620847914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,5120,3072,0.016184889607959323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,5120,2560,0.007983110845088959
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,3072,0.007930667036109501
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,3072,0.14584178394741482
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,12288,0.01719555589887831
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,5120,2048,0.006834666762087081
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,5120,2560,0.01477866702609592
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,2560,0.007838221887747446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,5120,2048,0.013772444592581855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,2048,0.14529066615634495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,5120,1536,0.005930666708283954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,5120,65536,0.10777599944008721
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,2048,0.007156444092591603
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,5120,1024,0.00452533322903845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,5120,1536,0.013186666700575085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,1536,0.006578666468461354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,5120,1024,0.012379555238617791
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,5120,768,0.0038942222793896994
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,1024,0.006280000011126201
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,4096,0.008626666333940294
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,5120,768,0.012064889073371887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,5120,3584,0.016920000314712524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,5120,512,0.00350222239891688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,768,0.14716355005900064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,768,0.006226666685607698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,5120,256,0.0031866667171319327
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,5120,512,0.012065778175989786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,512,0.00591111100382275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,512,0.1442906724082099
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,5120,128,0.0028088889602157804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,5120,256,0.011720889144473605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,256,0.005622222191757626
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,256,0.1416177749633789
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,2560,0.13963466220431858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,5120,64,0.0025191110455327565
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,5120,128,0.011075555450386472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,5120,32,0.002504000026318762
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,5120,128,0.00563822231358952
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,128,0.13164267275068495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,5120,64,0.011333333121405708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,5120,32,0.011750222080283694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,4096,65536,0.0859795543882582
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,4096,16384,0.026330666409598455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,4096,65536,0.07217955589294434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,4096,16384,0.026736888620588515
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,65536,0.06771910852856107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,4096,12288,0.020922667450375024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,16384,0.020463999774720933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,16384,0.14527644051445857
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,4096,10240,0.01850311126973894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,1024,0.14191555976867676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,4096,12288,0.022628444764349196
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,65536,0.17460178004370797
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,12288,0.016846223009957206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,12288,0.1401466661029392
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,4096,8192,0.01588444411754608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,4096,10240,0.020908445119857788
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,10240,0.015141333142916361
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,4096,7168,0.01462577780087789
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,10240,0.14207555188073054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,4096,8192,0.019919999771647982
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,8192,0.013036444783210754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,8192,0.14061866866217718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,4096,6144,0.013314666847387949
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,4096,7168,0.02033866610791948
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,7168,0.012115555504957834
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,4096,5120,0.011963555382357704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,7168,0.1467733383178711
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,6144,0.01069866700304879
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,4096,6144,0.019301333361201815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,6144,0.13934132787916395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,4096,4096,0.012746666868527731
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,4096,5120,0.018580444984965853
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,5120,0.009988444546858469
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,5120,0.14110310872395834
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,4096,3584,0.011680888632933298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,4096,4096,0.01688888834582435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,4096,0.14076711071862116
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,4096,3072,0.010317333042621613
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,4096,3584,0.01569955547650655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,3584,0.007985778152942657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,3584,0.13947555753919813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,4096,2560,0.0074728892909155945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,4096,3072,0.014815110299322339
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,3072,0.14021421803368464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,4096,2048,0.006099555641412735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,4096,2560,0.014117333624098035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,5120,1536,0.1444986661275228
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,4096,2048,0.013562666873137156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,2048,0.13962133725484213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,2048,0.007272000114123027
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,4096,1536,0.004867555366622077
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,1536,0.006615110983451207
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,4096,1024,0.004036444342798657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,4096,1024,0.012424888710180918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,1024,0.1384613381491767
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,4096,0.008346666892369589
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,1024,0.006228444476922353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,4096,768,0.0035235554807715942
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,4096,768,0.012376888758606382
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,3072,0.007872000336647034
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,768,0.1470444467332628
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,4096,512,0.0032106666929192017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,2560,0.14345778359307185
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,768,0.0059262220230367445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,4096,256,0.0032071111102898917
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,4096,512,0.011374221907721626
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,512,0.14396266142527261
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,512,0.005903111149867375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,4096,256,0.011697777443461947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,4096,128,0.0028071111689011254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,4096,1536,0.013109332985348172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,1536,0.14636977513631186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,4096,64,0.00282577777074443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,4096,128,0.011000889043013254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,4096,32,0.0028417776856157514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,128,0.1341208881802029
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,4096,64,0.011438222395049201
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3584,65536,0.07944355408350627
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,4096,32,0.011748444702890186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3584,65536,0.07117066780726115
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3584,16384,0.024358222881952923
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,65536,0.06672800249523587
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3584,16384,0.025987555583318073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,65536,0.17420621713002524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,16384,0.14108444584740534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,2560,0.0074311114019817775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3584,12288,0.019253333409627277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,16384,0.02014133334159851
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3584,10240,0.01683200067943997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3584,12288,0.02222488820552826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,12288,0.016695999436908297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3584,10240,0.0206675562593672
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,256,0.005717333406209946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3584,8192,0.014479999740918478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,4096,256,0.1450586716334025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,10240,0.14048088921440974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,10240,0.015044444137149386
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3584,7168,0.013648889131016202
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,4096,128,0.0058782220714622075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3584,8192,0.01994222237004174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,8192,0.013043555948469373
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3584,7168,0.01925511161486308
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,7168,0.011792889071835412
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,7168,0.14032533433702257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3584,5120,0.011668444507651858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,6144,0.14083645078870985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3584,5120,0.018240888913472492
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,5120,0.13999644915262857
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3584,4096,0.011729777687125735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,5120,0.009335111412737105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3584,4096,0.016332444217469957
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,4096,0.14210311571756998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3584,3584,0.011172444456153445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,4096,0.008392000363932716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,12288,0.14126755131615534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3584,3584,0.015796444482273526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3584,3072,0.009326222042242685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,3584,0.13981955581241184
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,3584,0.007947555846638149
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3584,2560,0.008990222381220924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3584,3072,0.014769777655601501
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,3072,0.007877333296669854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,3072,0.14497955640157065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,8192,0.14846577909257677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3584,2048,0.0058444444504049085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3584,2560,0.014157333307796054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,2560,0.14079911179012722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,2560,0.0076142218377855085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3584,6144,0.018458666072951425
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3584,1536,0.004575110971927643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,6144,0.010417777631017896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,2048,0.006983111302057902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,2048,0.14773066838582358
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3584,1024,0.003866666720973121
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3584,1536,0.01275644451379776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,1536,0.006783999916579988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3584,768,0.0038364442686239877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,1536,0.14395199881659612
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3584,1024,0.012387555506494312
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,1024,0.1456808911429511
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3584,512,0.0033093332830402586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3584,768,0.0123848890264829
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,768,0.14104978243509927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3584,256,0.0028737777223189673
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3584,512,0.011415999796655444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,512,0.005879999862776862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,512,0.140754673216078
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3584,128,0.0028008888992998335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3584,256,0.011584888729784222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,256,0.005879111174080107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,256,0.14707911014556885
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3584,64,0.0028391112056043413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3584,128,0.011013333168294696
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3584,32,0.002502222235004107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,128,0.005894222193294101
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,3584,128,0.13297422726949057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3584,64,0.01035822182893753
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3072,65536,0.06358310911390516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3584,32,0.012038222617573209
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3072,16384,0.020833777056799993
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3072,65536,0.06818044185638428
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,65536,0.06588088803821139
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3072,16384,0.025010666913456384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3584,2048,0.013628444737858243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,16384,0.14137866761949328
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,16384,0.019907555646366544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,65536,0.1735537714428372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3072,12288,0.021916444102923077
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,1024,0.0062773335311147906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,12288,0.14011111524369982
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,12288,0.016270221935378183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,3584,768,0.006117333140638139
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3072,8192,0.013078221844302284
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3072,10240,0.020259555843141343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,10240,0.14079022407531738
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,10240,0.014475555883513557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3072,7168,0.012152888708644442
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3072,8192,0.018681777848137748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,8192,0.01242844429281023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,8192,0.14006843831804064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3072,6144,0.011109333071443768
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3072,7168,0.019168888529141743
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,7168,0.011731555892361535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3072,6144,0.018396443790859647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3072,5120,0.01036088913679123
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,6144,0.14590488539801702
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3584,6144,0.012866666747464074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3072,5120,0.017248888810475666
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3072,4096,0.009626666704813639
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,5120,0.009360888765917884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3072,12288,0.017251556118329365
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3072,4096,0.015458666616015963
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3072,3584,0.009688888986905416
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,4096,0.14157156149546304
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,4096,0.008395555946562026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3072,10240,0.015327110886573792
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3072,3072,0.009318222602208456
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3072,3584,0.014512888259357877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,3584,0.00794577764140235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,3584,0.14192177189721003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3072,2560,0.006249777972698212
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3072,3072,0.014121777481502958
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,3072,0.007641777396202087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3072,2560,0.013775111072593264
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3072,2048,0.0052328887912962176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,2560,0.14486400286356607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,7168,0.14550756083594427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3072,2048,0.013356444736321768
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,2048,0.14397777451409235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3072,1536,0.004579555657174853
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,6144,0.010574222438865237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3072,1536,0.012764444781674279
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,1536,0.006631999793979857
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,1536,0.14115911059909397
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3072,1024,0.0038737778862317405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,5120,0.14642133977678087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3072,768,0.003535110917356279
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,1024,0.0063031112982167145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3072,1024,0.012392000191741519
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,1024,0.14127200179629856
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3072,768,0.011767110890812345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,768,0.006068444500366847
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,768,0.14566400316026476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3072,512,0.011391999820868174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3072,256,0.0031928889867332247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,512,0.005944889038801193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,3072,0.1398177809185452
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3072,256,0.011095110740926532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3072,128,0.0028408887899584244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,256,0.1450622214211358
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,2560,0.007287999822033777
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3072,128,0.010384000009960597
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3072,64,0.00249422217408816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,128,0.005617777920431561
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,128,0.1353164381451077
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3072,32,0.0028035555862718155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,2048,0.007221333682537079
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3072,64,0.01036977767944336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2560,65536,0.057982219590081104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,3072,32,0.011395555403497485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2560,16384,0.019495111372735765
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2560,65536,0.06537866592407227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,65536,0.0646373364660475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2560,16384,0.02445777753988902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,3072,512,0.0034933334423436057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2560,12288,0.015978667471143935
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,16384,0.14502044518788657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,65536,0.17356355985005698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,16384,0.019640889432695176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2560,10240,0.013758222262064615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2560,12288,0.02165422174665663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,12288,0.016153777639071148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,12288,0.14522843890719944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2560,10240,0.0199306673473782
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,3072,512,0.14124977588653564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,10240,0.14718222618103027
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2560,7168,0.011482666763994427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2560,8192,0.018552000323931377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,8192,0.011710222396585675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,8192,0.14008622699313694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2560,6144,0.01089422239197625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,3072,256,0.005718222094906702
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2560,7168,0.01870133313867781
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,7168,0.010684444672531553
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2560,5120,0.00980266680320104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2560,6144,0.0177875558535258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,6144,0.010319111247857412
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,6144,0.14192977216508654
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2560,5120,0.015824000040690105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2560,4096,0.008963555925422246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,5120,0.14041511217753092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,5120,0.009355555805895064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,4096,0.14119733704460993
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,4096,0.008286221987671321
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2560,4096,0.014807111687130399
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2560,3584,0.014114666316244336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,3584,0.008256889051861232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,3584,0.13994756009843615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2560,3072,0.01407911048995124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,3072,0.13774222797817653
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2560,2560,0.005928888916969299
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2560,8192,0.011959111524952782
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,3072,0.007975999679830339
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,10240,0.013946667313575745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2560,2048,0.0052391112678580815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2560,2560,0.013441777891582914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,2560,0.007636444436179266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,2560,0.1429671049118042
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2560,1536,0.004564444637960858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2560,2048,0.013050666285885705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,7168,0.14757955074310303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,2048,0.14080177413092718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2560,1024,0.0038453332251972626
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2560,1536,0.012554667062229581
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,1536,0.006565333240562015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,1536,0.14392622311909994
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2560,768,0.003519999898142285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2560,1024,0.012420444852775998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,1024,0.143504884507921
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2560,768,0.012381333443853589
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2560,3584,0.009277332988050249
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2560,512,0.0034231110993358824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,768,0.005943111247486538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2560,512,0.011702222128709158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2560,3072,0.00851733320289188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2560,256,0.0028560000161329904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,512,0.1415368848376804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2560,256,0.011724444727102915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,256,0.14045955075158015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2560,128,0.011071999867757162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,2048,0.007238222493065729
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,128,0.13037689526875815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2560,64,0.002798222212327851
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,128,0.005946666830115848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2560,32,0.00253866674999396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2560,64,0.011061333119869232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,1024,0.006276444428496891
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2048,65536,0.05091911223199633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2560,32,0.01145866678820716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,2560,768,0.14009066422780356
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2048,16384,0.016314667132165697
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2048,65536,0.06285511122809516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,512,0.005614222337802251
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,65536,0.06083466609319051
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2048,12288,0.01313511116637124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2048,16384,0.023949333363109167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,16384,0.14722843964894614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,2560,256,0.005906666732496685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2560,128,0.002838222309947014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2048,10240,0.011796444654464722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,12288,0.1393937799665663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2048,10240,0.01957066688272688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,10240,0.14217600557539198
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,10240,0.01316177762216992
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2048,8192,0.01111644423670239
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2048,8192,0.018107554978794523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,8192,0.011371555427710215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,8192,0.13976266649034289
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2048,7168,0.010688888529936472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2048,7168,0.016891555653678048
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,7168,0.010724444356229572
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,7168,0.13927911387549505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2048,6144,0.012367111113336353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2048,6144,0.01573777695496877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,6144,0.13963377475738525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2048,5120,0.011727111207114326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2048,5120,0.014871999621391296
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,5120,0.1421253283818563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,65536,0.17199733522203234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2048,4096,0.0103493332862854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,5120,0.009338666995366415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2048,12288,0.020949333906173706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,12288,0.014797333214018079
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,4096,0.14224089516533747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,4096,0.008341333104504479
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2048,3072,0.006903999795516332
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2048,3584,0.013526221944226159
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,3584,0.00794044468137953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,3584,0.1414639949798584
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2048,2560,0.005580444302823808
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2048,3072,0.013753777576817406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,3072,0.007640000018808577
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,3072,0.147599114312066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2048,2048,0.004943999979231092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,2560,0.1406924459669325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,2560,0.007287110719415877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2048,2048,0.01275022245115704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,2048,0.14007643858591715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,6144,0.010059555371602377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,2048,0.006959999601046245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2048,1536,0.0127368892232577
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,1536,0.146598219871521
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2048,1024,0.00396799999806616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,1536,0.00675733346078131
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,16384,0.01926933394538032
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2048,1024,0.012069332930776807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2048,4096,0.014073777529928418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2048,768,0.0035146665241983202
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,1024,0.006286222073766921
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2048,768,0.01203555530971951
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,768,0.005897777775923411
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2048,512,0.0034560000316964257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2048,512,0.011720000041855706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2048,2560,0.013417777915795645
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,512,0.13947289519839817
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2048,256,0.00286311118139161
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,512,0.0059004442559348206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2048,256,0.011350221931934357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,256,0.1415031088723077
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2048,128,0.0028231110837724474
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,256,0.005574222240183089
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2048,1536,0.004518222063779831
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2048,64,0.0028195555011431375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2048,128,0.011355555719799466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,2048,128,0.005613333235184352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,128,0.13156889544592962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2048,64,0.011014222270912595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1536,65536,0.036743111080593534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,2048,32,0.011711111499203576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1536,65536,0.06093600061204699
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2048,3584,0.009696000152164036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1536,16384,0.013438222308953604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,1024,0.1414817836549547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,65536,0.0596275528271993
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,65536,0.17248088783688018
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1536,12288,0.011519999967681037
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1536,16384,0.023353776997990076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,2048,768,0.14662844604916042
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,16384,0.1397226651509603
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1536,10240,0.010537777509954242
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1536,12288,0.02054666644997067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,12288,0.14717688825395372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1536,8192,0.009312888814343346
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1536,10240,0.01888799998495314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,10240,0.14605866538153753
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1536,8192,0.01755644381046295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1536,7168,0.009407111340098912
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,8192,0.14119111167060003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1536,7168,0.01651733285850949
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,7168,0.14119556215074328
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,2048,32,0.0025066667132907445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1536,6144,0.00907733374171787
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,7168,0.01073333372672399
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1536,5120,0.008658666577604082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,6144,0.009986666341622671
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,6144,0.13999999894036189
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1536,5120,0.014762666490342883
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1536,4096,0.007242666350470648
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,16384,0.018396443790859647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1536,4096,0.013980444934633044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,12288,0.0148062225845125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,4096,0.008286221987671321
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1536,3584,0.007296889192528195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,4096,0.14203288820054796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1536,3584,0.013792888985739814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,3584,0.00831911133395301
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,8192,0.011039999624093374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1536,3072,0.013404444687896304
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,3072,0.14557600021362305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1536,2560,0.005921777751710679
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,3072,0.007772444850868649
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1536,6144,0.015791111522250705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1536,2560,0.012867555850081973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,2560,0.14135821660359701
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,2560,0.007334222396214803
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,5120,0.009283555878533257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1536,2048,0.012899555265903473
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,2048,0.0069191112286514705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1536,1536,0.012431999875439538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1536,3072,0.006227555374304454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,1536,0.13932622803582087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1536,1024,0.0038862224254343244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,1536,0.006631999793979857
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1536,1024,0.01238044434123569
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,1024,0.14440977573394775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1536,768,0.0035013332962989807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,1024,0.006218666831652324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1536,2048,0.00591555568906996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1536,768,0.0120328888297081
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,5120,0.13987378279368082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1536,512,0.0031555555760860443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,768,0.005949333310127258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1536,1536,0.00424177779091729
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1536,512,0.011779555843936073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1536,256,0.003154666680428717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,512,0.14461333221859402
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,512,0.005662222289376789
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1536,128,0.0028364445186323593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,10240,0.012755555411179861
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,256,0.005751111027267244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,256,0.13996888531578913
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1536,128,0.010379555324713389
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1536,64,0.0028604444944196274
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1536,32,0.002532444480392668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,1536,128,0.005586666779385672
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,3584,0.14453155464596218
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,128,0.13192000654008654
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1024,65536,0.026399110754330952
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1536,64,0.01035733355416192
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1024,16384,0.014837332897716098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1536,32,0.011074666347768573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1024,65536,0.058360000451405845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,65536,0.05849777989917331
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,65536,0.17207911279466417
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1024,16384,0.02264533274703556
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1024,12288,0.011731555892361535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,16384,0.14155822330051
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,16384,0.01683999929163191
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1024,10240,0.010960000256697336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,12288,0.014098667436175875
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,12288,0.1486871110068427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1024,10240,0.017995556195576985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1024,8192,0.010686222049925061
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,10240,0.14156977335611978
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,10240,0.01273155543539259
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,768,0.14036711057027182
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1024,7168,0.010050666828950247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1024,8192,0.016576889488432143
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,8192,0.011367110742463006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,8192,0.14816178215874568
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1024,6144,0.009317333499590555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,1536,2048,0.14173155360751682
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1024,7168,0.016156444946924847
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,7168,0.010709332923094431
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1024,5120,0.008285333712895712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,7168,0.14551644855075413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1024,6144,0.015192889504962497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,6144,0.010035555395815108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1024,4096,0.007280000381999546
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1536,256,0.011749332977665795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,5120,0.009325332939624786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,5120,0.13904799355400935
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1024,3584,0.007306666837798224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1024,4096,0.014091556270917257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,4096,0.00832711077398724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,4096,0.14626399676005045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1024,3072,0.0062773335311147906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1024,3584,0.013801777528391944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,3584,0.00794666674402025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,3584,0.14492621686723497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1024,2560,0.005551110953092575
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1024,3072,0.01310755560795466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,3072,0.007595555649863348
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1024,2048,0.006596444381607904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1024,2560,0.013057777451144325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,2560,0.007300444775157505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,2560,0.14218578073713514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1024,1536,0.005592888842026393
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1024,2048,0.012890666723251343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1024,12288,0.018906666172875296
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,2048,0.006953777538405524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1024,1024,0.004612444589535396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1024,1536,0.012496888637542725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,1536,0.13991910881466335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1024,1024,0.01186044431395001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1024,768,0.0041964443193541635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,1024,0.14198399914635554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,1024,0.00620888877246115
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,768,0.005938666562239329
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1024,768,0.011720000041855706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,768,0.14145510726504856
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1024,512,0.003499555504984326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1024,5120,0.01494311127397749
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1024,256,0.0031644445326593188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1024,512,0.011381333072980246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,512,0.0058266665372583605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,512,0.13934489091237387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1024,128,0.0028702221396896574
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1024,256,0.011397333608733283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,256,0.14728977945115831
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1024,128,0.011345778074529437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,3072,0.14005333847469753
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1024,64,0.002528888897763358
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,1024,32,0.002842666581273079
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,128,0.13723822434743246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,128,0.005630222045713001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,768,65536,0.02195288903183407
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1024,64,0.011066666907734342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,1024,32,0.01204622205760744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,2048,0.14339021841684976
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,65536,0.058166219128502734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,65536,0.17217866579691568
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,1536,0.006636444479227066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,768,16384,0.021777777208222285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,768,12288,0.011123555401961008
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,16384,0.016864000095261466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,16384,0.14061689376831055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,768,12288,0.018935110833909776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,12288,0.1408026615778605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,768,10240,0.009721777505344814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,12288,0.014064888159434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,1024,6144,0.13984621895684138
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,768,8192,0.009660444325870937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,768,10240,0.017839110559887357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,10240,0.012768000364303589
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,768,8192,0.016364443633291457
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,768,7168,0.008871999879678091
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,8192,0.14067110750410292
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,8192,0.01128533316983117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,768,7168,0.01585155559910668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,7168,0.010593777729405297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,7168,0.1396311124165853
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,768,6144,0.00813155538505978
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,768,6144,0.015506666567590503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,6144,0.010034666293197209
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,768,5120,0.0075795559419526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,6144,0.14015555381774902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,768,65536,0.0566435522503323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,768,4096,0.007617777420414819
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,768,5120,0.014821334017647637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,5120,0.009334222310119206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,5120,0.14052355289459229
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,768,3584,0.007316444483068254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,768,4096,0.013591999808947245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,4096,0.00851822230550978
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,4096,0.1409697797563341
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,768,3072,0.0069306666652361555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,768,3584,0.013965333501497904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,3584,0.007986666427718269
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,3584,0.14661778344048396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,768,2560,0.006597333484225803
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,768,3072,0.01313244468635983
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,3072,0.1403262217839559
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,10240,0.14564800262451172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,768,2560,0.01275199982855055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,768,2048,0.00490399988161193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,2560,0.14085866345299616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,1024,256,0.005587555468082428
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,768,2048,0.0124195557501581
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,2048,0.14338666862911648
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,2048,0.0069271110826068455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,768,1536,0.012091555529170565
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,1536,0.006577777779764599
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,768,1024,0.0042168887125121225
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,1536,0.14549867312113443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,768,16384,0.01070755554570092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,1024,0.006230222268237009
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,768,1024,0.012111999922328524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,768,768,0.0038213332494099936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,1024,0.14813689390818277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,768,512,0.0034675554682811103
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,768,0.0059004442559348206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,768,0.14066221978929308
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,768,512,0.011386666860845355
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,768,256,0.003177777760558658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,512,0.005917333480384614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,512,0.14288443989223903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,768,256,0.011050666371981302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,256,0.005591111050711737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,768,128,0.002847111059559716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,3072,0.007631999750932057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,768,128,0.011414222419261932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,2560,0.007299555672539606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,128,0.13788533210754395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,768,64,0.002809777855873108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,768,32,0.0028364445186323593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,768,1536,0.004886222382386525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,512,65536,0.017627555463049147
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,768,64,0.010709332923094431
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,768,32,0.01202311118443807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,512,16384,0.011056000159846412
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,512,65536,0.05600444475809733
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,65536,0.17079467243618437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,512,16384,0.02162666618824005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,16384,0.13961511188083225
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,16384,0.016513778103722464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,768,768,0.012058667010731168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,512,12288,0.01901155544651879
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,12288,0.1478559970855713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,12288,0.013791999883121915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,512,10240,0.017465778523021273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,10240,0.14468888441721597
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,768,256,0.14140710565778944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,512,8192,0.00833333366447025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,10240,0.012611555556456247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,768,128,0.005595555735958948
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,512,8192,0.01649599936273363
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,512,7168,0.00832088871134652
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,8192,0.14348799652523464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,8192,0.01108266661564509
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,512,6144,0.007266666326257918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,512,7168,0.015792000624868605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,7168,0.01036533299419615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,7168,0.14171111583709717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,512,5120,0.007123555574152205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,512,6144,0.015126221709781222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,65536,0.05234488844871521
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,6144,0.14160888724856907
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,512,4096,0.007265778051482306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,5120,0.14421778255038792
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,512,12288,0.008998221821255153
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,512,4096,0.013944000005722046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,512,3584,0.006980444822046492
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,4096,0.008339555727110969
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,4096,0.14538844426472983
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,512,10240,0.008036444584528605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,512,3072,0.0069022224181228215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,512,3584,0.013459555804729462
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,3584,0.14180089367760554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,512,3072,0.013391111459996967
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,512,2560,0.006298666612969504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,3072,0.1422248946295844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,512,2560,0.01310488912794325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,2560,0.1413982179429796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,512,2048,0.006244444598754247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,2560,0.007304888632562425
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,512,2048,0.012729778057999082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,2048,0.1397955550087823
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,6144,0.010016000105275048
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,512,5120,0.014793778459231058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,512,1536,0.004576888763242298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,2048,0.0069297779765393995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,512,1024,0.0038355555799272326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,5120,0.009283555878533257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,512,1536,0.01241422196229299
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,1536,0.14028088251749674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,512,1024,0.012101333174440594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,512,768,0.0036391110883818734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,1024,0.14027822017669678
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,1024,0.006109333286682765
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,512,512,0.0035342222286595237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,512,768,0.011755555868148804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,3584,0.00793688909875022
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,768,0.1382533311843872
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,512,256,0.003136000078585413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,512,512,0.011733333269755045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,512,0.005889777921968036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,3072,0.007655111451943715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,512,128,0.002879111096262932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,512,256,0.011354666617181567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,256,0.005612444546487596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,256,0.14343644513024226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,512,64,0.0024968888610601425
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,512,128,0.011366222467687396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,128,0.13094045056237116
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,512,32,0.0028151110228565005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,128,0.005595555735958948
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,256,65536,0.01593777702914344
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,512,64,0.011015999648306104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,512,32,0.011683555940786997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,256,16384,0.008960888617568547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,256,65536,0.053971555497911244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,65536,0.05080799924002754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,65536,0.17279911041259766
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,256,12288,0.008271999657154083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,256,16384,0.021233777205149334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,16384,0.01681511104106903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,1536,0.006623999940024481
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,256,12288,0.019201777047581144
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,256,10240,0.007655999726719326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,12288,0.14025066958533391
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,12288,0.013753777576817406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,256,8192,0.007276444799370236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,256,10240,0.01789688898457421
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,10240,0.012730666332774691
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,512,768,0.005928888916969299
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,256,8192,0.0163102216190762
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,8192,0.1433324416478475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,512,512,0.14111110899183485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,256,7168,0.015799999237060547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,256,6144,0.006574222197135289
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,7168,0.1431208848953247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,7168,0.010513777534166971
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,256,6144,0.015126221709781222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,6144,0.009886222581068674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,256,5120,0.006288888967699475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,6144,0.14282577567630345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,5120,0.009292444421185387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,256,5120,0.014763555592960782
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,5120,0.1455377737681071
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,256,4096,0.005958222266700532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,256,4096,0.013807999591032663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,16384,0.14469510979122585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,256,3584,0.006262222097979651
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,256,3584,0.013804444836245643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,3584,0.14192177189721003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,10240,0.14284888903299967
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,3584,0.00795911086930169
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,256,7168,0.0069439998931354946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,256,2560,0.0063688887490166565
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,256,3072,0.013432000246312885
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,8192,0.01108888867828581
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,3072,0.13998311095767552
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,3072,0.007626666790909237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,256,2048,0.006114666660626729
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,2560,0.0075457774930530125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,2560,0.14076799816555446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,256,2048,0.01240888900227017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,256,1536,0.004549333204825719
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,2048,0.14069778389400905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,2048,0.00692622239391009
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,256,1536,0.012059555285506778
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,1536,0.006576889091067844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,256,1024,0.003863111138343811
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,256,1024,0.012102222277058495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,1024,0.14760977692074245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,4096,0.008215111162927415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,256,768,0.0035297779573334586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,4096,0.1456835534837511
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,1024,0.006278222219811545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,256,512,0.003168888803985384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,256,768,0.012077333198653327
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,768,0.005902222047249476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,768,0.1459840006298489
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,256,3072,0.005919999960396025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,256,256,0.0028408887899584244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,256,512,0.011709333293967776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,512,0.005888888819350137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,512,0.14205866389804417
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,256,128,0.0028657778683635923
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,256,256,0.0114702218108707
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,256,0.005559111220969095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,256,0.14645244015587702
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,256,128,0.011128000087208219
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,256,128,0.005577777822812398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,256,32,0.002521777732504739
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,256,2560,0.01312800000111262
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,128,65536,0.0144177774588267
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,256,32,0.011448000040319232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,128,65536,0.05204711026615567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,128,16384,0.008261333737108443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,65536,0.17170311344994438
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,65536,0.05024888780381945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,1536,0.13988267050849065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,128,12288,0.007644444704055786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,16384,0.14223644468519422
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,16384,0.016895999511082966
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,128,12288,0.01888088881969452
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,128,10240,0.007258666886223688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,12288,0.013935110635227628
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,128,10240,0.017877333694034152
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,10240,0.14598133828904894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,128,8192,0.0069617778062820435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,10240,0.012440000143316058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,256,64,0.00253866674999396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,128,8192,0.016508445143699646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,128,7168,0.00665244460105896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,256,128,0.13585333029429117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,8192,0.14127111434936523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,256,64,0.010036444498433007
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,128,6144,0.006249777972698212
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,128,7168,0.015814221567577787
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,7168,0.01036622209681405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,7168,0.14625333415137395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,128,5120,0.006251555350091722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,128,6144,0.014944000376595391
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,6144,0.00996088898844189
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,6144,0.1464080015818278
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,128,16384,0.02129777769247691
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,128,5120,0.014446222119861178
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,128,4096,0.005913777897755305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,5120,0.14827556080288357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,128,3584,0.006287999865081575
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,4096,0.0083155557513237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,4096,0.14407377772861057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,12288,0.1428604390886095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,128,3072,0.006098666538794835
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,3584,0.007931555310885111
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,128,3072,0.01370844410525428
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,3072,0.14404000176323786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,3072,0.007629333270920648
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,128,2560,0.006568000134494569
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,128,2560,0.013182222015327878
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,8192,0.011019555230935415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,2560,0.1447626617219713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,128,2048,0.005945777727497949
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,2560,0.007302222152551015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,128,2048,0.012723555167516073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,128,1536,0.004516444272465176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,2048,0.14237956206003824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,128,1536,0.012381333443853589
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,1536,0.14188267125023737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,128,1024,0.0038399998512532976
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,128,4096,0.013763555222087435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,1536,0.006634666687912411
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,128,768,0.003549333247873518
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,128,3584,0.013152889079517789
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,128,1024,0.012039110892348819
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,1024,0.14103466934627956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,3584,0.1462062199910482
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,128,768,0.011761777930789523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,768,0.006057777752478917
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,768,0.14582044548458523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,128,256,0.002895111011134254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,128,512,0.011735111474990845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,512,0.005952888892756567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,512,0.14383467038472494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,128,256,0.01071200023094813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,128,128,0.0026871110830042097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,256,0.005585777676767773
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,128,128,0.009852444132169088
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,128,0.14156799846225315
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,128,64,0.002510222295920054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,2048,0.006934222247865465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,128,32,0.002489777695801523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,128,64,0.011003555523024665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,5120,0.009187555147541894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,64,65536,0.011065777805116443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,128,32,0.012078222301271228
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,64,65536,0.05133777856826782
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,64,16384,0.021623111433453027
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,64,12288,0.006896889044178857
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,1024,0.006068444500366847
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,64,10240,0.006266666783226862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,128,512,0.0032115555885765287
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,64,8192,0.006616000086069107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,64,7168,0.00629066675901413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,64,12288,0.0192275560564465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,64,10240,0.017847999930381775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,64,6144,0.00627377794848548
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,64,8192,0.01647733317481147
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,64,5120,0.005881777654091517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,64,4096,0.0058871110280354815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,64,7168,0.0158595558669832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,64,3584,0.006258666515350342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,64,5120,0.014483556151390076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,16,128,256,0.1426524453692966
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,64,3072,0.005944889038801193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,64,4096,0.014147554834683737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,64,2560,0.006280000011126201
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,64,3584,0.013759999639458127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,64,2048,0.0058497778243488735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,64,1536,0.004550222307443619
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,64,2560,0.013406222065289816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,64,1024,0.0035413333939181436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,64,2048,0.012810666528013019
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,16,128,128,0.005583999885453119
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,64,1536,0.012435555458068848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,64,512,0.003175111073586676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,64,1024,0.012436444560686747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,64,256,0.002801777794957161
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,64,768,0.011409777734014722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,64,128,0.002502222235004107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,64,16384,0.007132444116804335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,64,64,0.002479111154874166
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,64,256,0.011415111521879831
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,64,128,0.010010666317409938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,64,64,0.011050666371981302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,32,65536,0.009490666290124258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,64,32,0.010422222316265106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,32,16384,0.006292444550328785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,32,65536,0.05124977893299527
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,32,12288,0.006245333287451003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,64,6144,0.01516888870133294
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,32,10240,0.006222222414281633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,32,12288,0.01886044442653656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,32,8192,0.005912888795137405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,32,7168,0.006583111153708563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,32,10240,0.017835555805100333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,32,6144,0.006266666783226862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,32,8192,0.016422222057978313
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,32,5120,0.006213333457708359
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,32,7168,0.015493333339691162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,32,4096,0.005892444401979446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,32,6144,0.0151555554734336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,32,3584,0.00591466658645206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,32,5120,0.014418666561444601
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,64,512,0.011713777979214987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,32,3072,0.005872888697518243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,32,4096,0.013759999639458127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,32,2560,0.006235555642180973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,32,2048,0.005994666781690385
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,64,32,0.002534222271707323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,32,3072,0.013391111459996967
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,32,1536,0.004516444272465176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,32,2560,0.013063111239009433
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,32,1024,0.003559111307064692
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,32,2048,0.012698666916953193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,32,1536,0.012448000411192576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,32,1024,0.012093333734406365
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,32,512,0.003141333245568805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,32,256,0.002887111157178879
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,32,512,0.011717333561844297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,32,256,0.011711111499203576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,32,128,0.011381333072980246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,64,3072,0.01367911116944419
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,32,64,0.002489777695801523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,32,32,0.0028275555620590844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,64,768,0.0034871109657817413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,32,64,0.011379555695586734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,32,32,0.011018666956159802
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,32,3584,0.013759999639458127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,65536,16384,0.17069599363538954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,65536,16384,0.34249154726664227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,32,768,0.003391111062632667
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,65536,12288,0.24606667624579537
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,32,768,0.012057777908113269
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,65536,12288,0.15069954925113255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,16,32,16384,0.020638222495714825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,16,32,128,0.0028080000645584534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,16384,0.1389413277308146
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,65536,10240,0.20617511537339953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,65536,10240,0.133607996834649
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,12288,0.10445066955354478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,65536,8192,0.1546364492840237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,10240,0.09185244639714558
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,65536,8192,0.11454310682084824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,65536,7168,0.13331111272176108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,65536,7168,0.10007199976179336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,8192,0.07022044393751356
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,16384,0.21352622244093153
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,65536,6144,0.11473777559068467
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,10240,0.1514835490120782
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,7168,0.06334488921695285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,65536,6144,0.08955822388331096
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,8192,0.14134132862091064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,65536,5120,0.09667199850082397
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,65536,5120,0.08041333489947848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,7168,0.14135732915666369
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,65536,4096,0.08057689004474215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,6144,0.14106310738457575
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,5120,0.04726755619049072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,65536,4096,0.06553510824839275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,5120,0.13854311572180852
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,65536,3584,0.06896177927652995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,65536,3584,0.05934311283959282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,4096,0.03868088788456387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,4096,0.14191555976867676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,65536,3072,0.059679110844930015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,3584,0.0351582235760159
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,12288,0.17094133959876168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,65536,2560,0.05084177851676941
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,3584,0.14044088787502712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,3072,0.03141599893569946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,65536,2560,0.04821688930193583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,65536,2048,0.04176177912288242
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,3072,0.1413360039393107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,2560,0.027160889572567407
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,2560,0.14778578281402588
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,65536,1536,0.03231111168861389
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,6144,0.05548355645603604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,2048,0.02325244413481818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,65536,1024,0.023323555787404377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,2048,0.13894577821095785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,65536,1536,0.03736799955368042
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,1536,0.018970666660202872
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,1536,0.14005066288842097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,65536,768,0.01868088874551985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,65536,1024,0.023476445012622412
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,1024,0.0173653331067827
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,65536,512,0.013458666702111563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,65536,768,0.022614222433831956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,768,0.14451289176940918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,768,0.015145777000321282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,65536,512,0.018609777092933655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,65536,256,0.008323555191357931
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,512,0.13850843906402588
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,512,0.012929778132173749
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,65536,128,0.006280000011126201
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,65536,3072,0.05134044422043694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,65536,256,0.015302222636010913
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,256,0.010444444086816577
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,65536,64,0.005572444448868434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,65536,128,0.014153778553009033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,65536,128,0.010355555348926121
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,65536,64,0.016231111354298063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,65536,32,0.015178667174445258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,65536,2048,0.04115466607941522
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,16384,65536,0.32220623228285045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,16384,65536,0.18316800064510771
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,1024,0.13994667265150282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,16384,16384,0.0868684450785319
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,16384,16384,0.0566577778922187
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,16384,0.044450667169359
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,256,0.13915022214253744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,16384,0.13988355795542398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,65536,0.17850844065348306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,65536,128,0.1323431067996555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,16384,12288,0.0542791121535831
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,16384,10240,0.054903109868367515
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,12288,0.03892799880769517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,12288,0.13896089129977757
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,10240,0.029934220843844946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,16384,8192,0.04289955563015408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,10240,0.1466568840874566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,16384,8192,0.0440631111462911
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,8192,0.0247662216424942
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,16384,7168,0.037760890192455716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,8192,0.13933332761128744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,16384,7168,0.04003377755482992
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,65536,0.22454844580756295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,16384,6144,0.03218400147226121
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,7168,0.022701332966486614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,16384,5120,0.028445333242416382
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,7168,0.140374223391215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,6144,0.020072888996866014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,16384,12288,0.06751822100745307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,6144,0.14223821957906088
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,16384,5120,0.03154044349988302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,5120,0.13819555441538492
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,16384,4096,0.023697778582572937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,65536,32,0.005555555638339784
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,16384,4096,0.027021333575248718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,16384,3584,0.020397333635224234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,4096,0.015393777026070489
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,4096,0.13839466041988796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,16384,3072,0.0181697772608863
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,16384,3584,0.024382222029897902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,16384,10240,0.05607110924190945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,3584,0.14484443929460314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,3584,0.014311111635631986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,16384,2560,0.015506666567590503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,16384,3072,0.02334311107794444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,3072,0.013261333107948303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,3072,0.14797155062357584
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,16384,2560,0.021578666236665513
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,16384,2048,0.01309866706530253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,2560,0.14517333772447374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,16384,1536,0.01036533299419615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,2048,0.010946667028797997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,2048,0.14237156179216173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,16384,6144,0.035651554663976036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,16384,1536,0.017473777135213215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,1536,0.009387555221716562
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,1536,0.1403448846605089
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,16384,1024,0.008667555948098501
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,16384,768,0.0069164443347189165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,16384,1024,0.014880888991885714
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,1024,0.0086666668454806
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,1024,0.141595549053616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,16384,512,0.005408000200986862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,5120,0.018192888961897958
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,768,0.007983999947706858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,768,0.1451902257071601
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,16384,256,0.0034871109657817413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,16384,512,0.013350221845838757
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,512,0.007259555160999298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,512,0.14590666029188368
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,16384,256,0.012090666426552666
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,256,0.14124533865186903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,256,0.006640888750553131
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,16384,128,0.011365333365069495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,128,0.006636444479227066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,16384,64,0.0028071111689011254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,16384,32,0.0028640000770489373
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,16384,2560,0.0120666664507654
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,16384,64,0.012721777790122561
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,16384,32,0.01203644441233741
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,16384,2048,0.01883288886811998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,12288,65536,0.2444417741563585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,12288,65536,0.1583226654264662
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,12288,16384,0.06672710842556424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,16384,768,0.014472888575659858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,12288,16384,0.04968533251020643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,65536,0.14400622579786512
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,12288,12288,0.051947553952534996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,16384,0.04658133453792996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,12288,12288,0.04847200049294365
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,16384,0.13991999626159668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,12288,10240,0.04432622260517544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,12288,0.03131555517514547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,12288,0.14108178350660536
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,12288,10240,0.0436737769179874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,16384,128,0.13507200611962214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,10240,0.026663111315833196
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,10240,0.1402133305867513
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,12288,7168,0.0326257778538598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,8192,0.021931555536058214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,12288,8192,0.038727111286587186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,8192,0.1389146645863851
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,12288,6144,0.028463999430338543
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,65536,0.17817778057522246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,7168,0.020049777295854356
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,7168,0.145395557085673
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,12288,6144,0.0326257778538598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,12288,5120,0.025115556187099878
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,6144,0.14056889216105142
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,6144,0.018422222799725003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,12288,4096,0.022457778453826904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,5120,0.0162773331006368
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,5120,0.13900089263916016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,12288,3584,0.01947377787695991
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,12288,4096,0.025401777691311304
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,4096,0.014360000689824423
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,4096,0.13976621627807617
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,12288,3072,0.0174346665541331
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,12288,3584,0.02278933260175917
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,3584,0.013270222478442721
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,16384,128,0.0031928889867332247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,3584,0.14563466442955866
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,12288,2560,0.01586844523747762
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,12288,3072,0.021918222308158875
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,3072,0.012144889268610211
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,3072,0.13913333415985107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,12288,8192,0.03735911183887058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,12288,2560,0.02010222276051839
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,12288,2048,0.013426666458447775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,2560,0.011128889189826118
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,2560,0.14520800113677979
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,12288,2048,0.01826844447188907
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,2048,0.009699555734793345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,2048,0.14013155301411948
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,12288,1024,0.006919999917348226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,12288,1536,0.017490666773584154
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,1536,0.1386879947450426
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,12288,7168,0.035346666971842446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,12288,768,0.0058604445722368025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,1024,0.007992000215583378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,1024,0.14459466934204102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,12288,5120,0.02872177627351549
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,12288,768,0.014159111513031853
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,12288,512,0.003879111260175705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,768,0.007812444534566667
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,768,0.13881511158413357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,12288,256,0.0032044444233179092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,12288,512,0.01310488912794325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,512,0.007122666471534305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,512,0.14479733837975398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,12288,256,0.012093333734406365
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,12288,128,0.0028613333900769553
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,256,0.13915467262268066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,12288,128,0.011408000356621213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,12288,64,0.0028560000161329904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,12288,128,0.13649066289265951
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,12288,32,0.0028728888266616394
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,12288,64,0.012387555506494312
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,12288,1536,0.008760888543393876
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,12288,32,0.011074666347768573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,1536,0.008610666626029545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,10240,65536,0.20562399758232963
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,12288,1024,0.01422933406300015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,10240,65536,0.15211733182271323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,10240,16384,0.05859733290142483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,10240,16384,0.04725955592261421
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,10240,12288,0.044171555174721606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,16384,0.04363644454214308
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,65536,0.13136088848114014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,16384,0.1408240000406901
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,10240,12288,0.04390311241149902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,12288,0.03026755650838216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,256,0.006653333289755716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,12288,0.13902666833665636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,10240,10240,0.038655112187067665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,10240,8192,0.032079110542933144
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,12288,128,0.006513777706358168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,10240,0.025447999437650044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,10240,0.14271733495924208
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,65536,0.15923199388715956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,10240,7168,0.028833776712417603
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,8192,0.021574222379260596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,8192,0.1377520031399197
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,10240,6144,0.025951999757024977
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,10240,7168,0.03163999981350369
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,7168,0.019817777805858187
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,10240,5120,0.022429333792792425
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,7168,0.14534933037228054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,10240,6144,0.027448000179396734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,6144,0.017649778061442904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,6144,0.13880088594224718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,10240,5120,0.025027554896142747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,5120,0.016057777735922072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,10240,4096,0.022727110319667395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,4096,0.14861955907609728
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,4096,0.014005333185195923
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,10240,3584,0.017260443833139207
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,10240,10240,0.03788266579310099
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,10240,3584,0.021380444367726643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,3584,0.013152889079517789
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,3584,0.13873244656456843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,10240,3072,0.015595555305480957
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,10240,2560,0.014370666609870063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,10240,3072,0.020492444435755413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,3072,0.13987733258141413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,10240,8192,0.033837334977255926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,2560,0.010808000134097205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,2560,0.14425955878363714
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,10240,2048,0.010607110957304636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,10240,2048,0.017808000246683758
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,2048,0.009663110805882348
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,2048,0.14527111583285862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,10240,1536,0.015572445260153877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,10240,4096,0.019239111079110038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,1536,0.143103109465705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,10240,1024,0.006561777657932705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,5120,0.14538844426472983
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,1536,0.008637333081828224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,10240,768,0.004903111192915174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,10240,1024,0.013616888887352414
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,1024,0.14636622534857854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,10240,512,0.0034871109657817413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,10240,768,0.013417777915795645
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,768,0.007639110916190677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,768,0.14617244402567545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,10240,256,0.0031991110493739447
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,10240,512,0.012216889195972018
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,512,0.006953777538405524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,512,0.14109599590301514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,10240,256,0.011008000208271874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,10240,128,0.002842666581273079
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,256,0.14393333594004312
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,10240,2560,0.018544000056054857
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,10240,128,0.010464888479974536
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,10240,128,0.13577600320180258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,10240,64,0.0028435554769304064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,10240,32,0.002850666642189026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,128,0.006344888773229387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,10240,1536,0.008978666530715095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,10240,64,0.012081777883900536
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,10240,32,0.011044444309340583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,8192,65536,0.16955555809868705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,8192,65536,0.09810488753848606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,1024,0.007662222617202335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,65536,0.07986844248241849
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,8192,16384,0.03579377796914842
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,8192,12288,0.03283199999067519
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,16384,0.024010666542583044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,16384,0.14207200209299722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,3072,0.012076444096035428
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,8192,12288,0.03529511226548089
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,10240,256,0.006623999940024481
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,8192,10240,0.027822222974565294
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,12288,0.14797600110371908
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,65536,0.15026044845581055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,12288,0.01980266637272305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,8192,10240,0.032574223147498235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,8192,8192,0.023939554889996845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,10240,0.01724177764521705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,10240,0.14255289236704508
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,8192,0.015088889333936902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,8192,8192,0.026757333013746474
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,8192,6144,0.018056000272432964
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,7168,0.014053333136770459
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,8192,16384,0.04531466629770067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,8192,6144,0.022806222240130108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,8192,5120,0.015663999650213454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,6144,0.14141066869099936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,6144,0.012914666699038612
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,8192,5120,0.02095199955834283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,5120,0.1393644412358602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,5120,0.012055111428101858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,8192,4096,0.019754666421148512
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,4096,0.1391057835684882
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,8192,3584,0.011716444459226398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,4096,0.011000889043013254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,8192,3584,0.01903733279969957
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,8192,7168,0.020617778102556866
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,3584,0.13804977469974095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,8192,3072,0.011031111081441244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,8192,0.14263555738661024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,3584,0.009647111097971598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,8192,2560,0.009467555416954888
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,8192,7168,0.024657777614063684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,8192,3072,0.01822933389080895
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,7168,0.1413386662801107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,8192,2048,0.008539555801285638
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,8192,2560,0.01668177710639106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,2560,0.008299555215570662
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,2560,0.1417751047346327
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,8192,1536,0.007224889265166388
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,8192,2048,0.0155102229780621
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,2048,0.007754666937722101
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,8192,1024,0.0055440002017551
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,8192,1536,0.0147733340660731
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,1536,0.13843377431233725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,8192,1024,0.013062222136391534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,8192,4096,0.013552000125249227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,1024,0.006595555692911148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,1024,0.1420960028966268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,768,0.006611555400821898
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,768,0.13926488823360866
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,8192,512,0.01200799975130293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,512,0.13915377193027073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,512,0.006247111078765657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,8192,256,0.0033395555284288195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,8192,256,0.01142133358452055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,2048,0.14461867014567056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,1536,0.007272889216740926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,8192,128,0.0028719999310043124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,256,0.1391671101252238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,8192,768,0.003703111161788305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,8192,128,0.011414222419261932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,128,0.13082488377888998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,8192,64,0.002540444541308615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,8192,768,0.01273333364062839
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,8192,32,0.0028773333049482773
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,8192,64,0.012034667034943899
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,8192,32,0.011372444530328115
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,7168,65536,0.1499404509862264
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,7168,65536,0.09268089135487874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,7168,16384,0.041366222831938006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,3072,0.009692444569534725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,65536,0.07465155257119073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,7168,16384,0.03352888756328159
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,7168,12288,0.03310577736960517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,16384,0.022692445251676772
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,16384,0.14639466338687473
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,256,0.005934222290913264
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,7168,12288,0.03386488887998793
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,65536,0.1513599952061971
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,7168,10240,0.0284826656182607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,12288,0.14172444078657362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,12288,0.018732445107565988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,8192,128,0.005952888892756567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,7168,8192,0.024273778001467388
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,10240,0.016523554921150208
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,10240,0.14223821957906088
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,7168,7168,0.018773333893881906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,8192,0.014550222290886773
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,8192,0.14419466919369167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,7168,7168,0.024499555428822834
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,7168,0.1392542256249322
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,7168,6144,0.01667022208372752
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,8192,3072,0.13777511649661595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,7168,6144,0.022327999273935955
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,6144,0.13946133189731175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,7168,5120,0.01456622282663981
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,6144,0.01239822225438224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,7168,5120,0.02083822256988949
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,5120,0.14078044891357422
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,7168,4096,0.012741333080662621
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,7168,4096,0.019268444842762418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,4096,0.13962577448950872
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,7168,3584,0.011414222419261932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,4096,0.009872000250551436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,8192,512,0.0034968890249729156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,7168,3584,0.018034666776657104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,7168,8192,0.026068444053332012
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,3584,0.009023110899660323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,7168,3072,0.010182222558392419
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,3072,0.008375999828179678
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,7168,3072,0.01719911065366533
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,3072,0.14165866374969482
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,7168,2560,0.008956444760163626
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,7168,0.013427555561065674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,7168,2048,0.007924444145626491
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,7168,2560,0.01640977793269687
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,2560,0.007789333661397298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,2560,0.14223110675811768
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,7168,1536,0.006567111031876669
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,7168,2048,0.014807111687130399
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,2048,0.007296000089910295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,2048,0.13950310813056097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,7168,1024,0.004559999952713649
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,7168,1536,0.014482667048772177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,1536,0.006945778098371293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,5120,0.01168444421556261
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,1536,0.14458221859402126
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,7168,768,0.0038382220599386427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,7168,1024,0.012714666624863943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,1024,0.006559111177921295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,7168,512,0.003206222214632564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,7168,768,0.012512889173295764
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,768,0.006272888845867581
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,768,0.1386924452251858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,7168,256,0.0031786666562159858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,7168,10240,0.0314008891582489
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,512,0.005924444645643234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,512,0.14351467291514078
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,7168,128,0.0028328889360030494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,7168,256,0.011416888899273343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,3584,0.14245066377851698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,256,0.005921777751710679
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,256,0.14518222543928358
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,7168,64,0.002562666725781229
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,7168,128,0.011382222175598145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,128,0.1318275531133016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,7168,128,0.005681777579916849
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,7168,64,0.011767110890812345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,7168,32,0.011061333119869232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,6144,65536,0.1220124430126614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,6144,16384,0.03538844320509169
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,65536,0.07219111257129245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,6144,16384,0.03154311246342129
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,7168,1024,0.14182399378882513
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,16384,0.02186311119132572
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,6144,12288,0.028752889898088243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,16384,0.14499466949039036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,65536,0.15040444003211126
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,6144,12288,0.030863109562132094
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,6144,10240,0.024484443995687697
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,7168,512,0.012019555601808759
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,12288,0.01792266633775499
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,12288,0.1423528856701321
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,6144,8192,0.021418665846188862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,10240,0.016050666570663452
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,6144,10240,0.02772355576356252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,6144,7168,0.01889333294497596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,10240,0.14471556080712214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,7168,32,0.0028044444819291434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,6144,8192,0.025225778420766194
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,8192,0.01406044430202908
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,6144,6144,0.016785777277416654
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,8192,0.1473884450064765
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,6144,7168,0.022914666268560622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,7168,0.013031110995345645
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,6144,5120,0.015415999624464246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,7168,0.1397306654188368
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,6144,6144,0.022261333134439256
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,6144,0.012303111453851065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,6144,4096,0.011727111207114326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,6144,65536,0.08587911393907334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,6144,5120,0.020336000455750358
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,5120,0.01107911103301578
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,6144,3584,0.010404444403118556
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,6144,4096,0.018555555078718398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,4096,0.00960533320903778
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,6144,3072,0.009535999761687385
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,3584,0.00832266691658232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,3584,0.14083288775549993
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,6144,3072,0.01683022247420417
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,6144,2560,0.008609777523411645
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,3072,0.14055111673143175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,6144,2560,0.015460444821251763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,2560,0.13989688290490046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,6144,2048,0.007174222005738153
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,2560,0.007956444389290279
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,6144,2048,0.014819555812411837
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,2048,0.1385235521528456
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,6144,1536,0.006041777630647023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,2048,0.007292444507280986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,6144,0.14409777853224012
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,6144,1536,0.013464888764752282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,6144,1024,0.004548444516128964
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,1536,0.14050489001803929
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,5120,0.14431999789343938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,1536,0.0069511110583941145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,4096,0.1421146657731798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,6144,768,0.003858666867017746
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,6144,3584,0.017486222916179232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,6144,1024,0.012738666600651212
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,1024,0.006281777802440856
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,1024,0.14040622446272108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,6144,512,0.0032248888164758682
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,6144,768,0.012727110750145383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,768,0.006240889016124938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,6144,256,0.003162666741344664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,6144,512,0.011718221836619906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,512,0.005967999911970562
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,512,0.13901600572797987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,6144,128,0.0028213332924577924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,6144,256,0.011364444262451597
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,256,0.005603555589914322
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,6144,128,0.011066666907734342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,6144,64,0.002797333316670524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,128,0.13271733125050864
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,6144,32,0.0028577778074476454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,128,0.005899555567238066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,6144,64,0.012083555261294046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,6144,32,0.010674667027261523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,5120,65536,0.10629244645436604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,5120,65536,0.07675288783179389
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,5120,16384,0.03176533182462057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,5120,16384,0.029719998439153034
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,65536,0.06953599717881945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,5120,12288,0.025861332813898723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,16384,0.1460382276111179
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,16384,0.02127911150455475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,5120,10240,0.022108443909221228
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,65536,0.15007111761305067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,12288,0.017475555340449016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,768,0.14447200298309326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,5120,10240,0.02676711148685879
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,6144,3072,0.008622222476535374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,5120,8192,0.020022221737437777
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,10240,0.0155404441886478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,6144,256,0.14097244209713405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,5120,7168,0.017136888371573556
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,8192,0.013705777625242868
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,8192,0.1401724418004354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,5120,6144,0.015453333655993143
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,5120,7168,0.023336889015303716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,7168,0.012748444245921241
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,7168,0.13922488689422607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,5120,5120,0.013763555222087435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,5120,6144,0.02113955550723606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,6144,0.011598221957683563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,6144,0.14109867148929173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,5120,4096,0.01126399967405531
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,5120,5120,0.019875556230545044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,5120,0.13815555307600233
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,5120,0.01036622209681405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,5120,3584,0.009656888743241629
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,5120,4096,0.018072888255119324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,4096,0.13954667250315347
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,4096,0.008967999782827165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,5120,3072,0.008847999903890822
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,5120,3584,0.01720355616675483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,5120,12288,0.02950577934583028
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,3584,0.008386666576067606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,3584,0.14209511544969347
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,5120,2560,0.007876444194051955
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,12288,0.1451688872443305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,3072,0.007920888562997183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,3072,0.1453875568177965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,5120,2560,0.014997333288192749
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,2560,0.13935555352105036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,2560,0.007613333563009898
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,10240,0.14019644260406494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,5120,1536,0.005600000007285013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,5120,2048,0.014107555150985718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,2048,0.1415839989980062
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,2048,0.006979555719428592
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,5120,8192,0.024116444918844435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,5120,1024,0.0038382220599386427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,5120,1536,0.013393777940008374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,1536,0.006619555668698416
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,1536,0.14462578296661377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,5120,768,0.003548444559176763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,5120,1024,0.012071111136012606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,1024,0.14478933811187744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,1024,0.006222222414281633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,5120,512,0.0032702222880389956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,5120,768,0.012442666623327466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,768,0.006051555689838197
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,768,0.14070399602254233
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,5120,256,0.002855111120475663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,5120,512,0.012089778151777057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,512,0.005967999911970562
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,512,0.14394844902886286
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,5120,128,0.0028533333291610083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,5120,256,0.011387555963463254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,256,0.14662755860222712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,5120,64,0.0028311111446883944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,5120,128,0.011025778121418424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,128,0.005774222314357758
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,5120,128,0.13628621896107992
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,5120,32,0.002814222127199173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,5120,64,0.011048888994587792
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,5120,32,0.011406222151385413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,5120,3072,0.015840000576443143
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,4096,16384,0.02624888883696662
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,5120,2048,0.006904888898134232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,65536,0.06732889016469319
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,4096,16384,0.025980444418059453
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,4096,12288,0.0207022229830424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,16384,0.020473778247833252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,65536,0.14977422025468615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,4096,12288,0.026584888497988384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,4096,10240,0.018325333793958027
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,12288,0.0166631109184689
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,12288,0.1425573296017117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,4096,10240,0.024663110574086506
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,10240,0.015086222026083203
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,10240,0.14675466219584146
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,5120,256,0.005906666732496685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,4096,8192,0.022662222385406494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,8192,0.012990222209029727
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,4096,65536,0.09082311391830444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,4096,7168,0.021576889687114294
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,4096,65536,0.070942223072052
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,4096,6144,0.013434666726324292
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,7168,0.14361955059899226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,7168,0.011733333269755045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,16384,0.1402319934633043
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,4096,6144,0.019827556278970506
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,6144,0.010716444088353051
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,6144,0.14379199345906576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,4096,5120,0.011760888828171624
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,4096,4096,0.010675555302037133
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,4096,5120,0.018583999739752877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,5120,0.010016000105275048
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,4096,4096,0.01716444392999013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,4096,3584,0.010764444039927589
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,4096,0.008656000097592672
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,4096,0.14543288283877903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,4096,8192,0.015836444165971544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,4096,3584,0.016186666157510545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,3584,0.008300444318188561
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,3584,0.14442933930291071
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,4096,3072,0.014866666661368476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,4096,7168,0.01441422270403968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,4096,2560,0.008617777791288164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,3072,0.008003555238246918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,3072,0.14528711636861166
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,4096,2048,0.007376000285148621
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,4096,2560,0.01442488862408532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,2560,0.0074728892909155945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,2560,0.17241244845920137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,4096,1536,0.006016888966162999
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,4096,2048,0.013413333230548434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,2048,0.007069333559936947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,2048,0.14216532972123888
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,4096,1024,0.00519822206762102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,4096,1536,0.013089777694808112
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,1536,0.006948444578382704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,1536,0.14402666356828478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,4096,768,0.004522666749027041
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,4096,1024,0.012423999607563019
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,1024,0.006262222097979651
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,1024,0.13978400495317247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,5120,0.14127555158403185
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,4096,512,0.00397866674595409
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,4096,768,0.012105777859687805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,768,0.141903109020657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,768,0.006269333263238271
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,4096,256,0.0031306667046414483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,4096,512,0.012078222301271228
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,512,0.1385093397564358
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,4096,128,0.0028444443725877335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,4096,256,0.011377777490350934
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,256,0.00589155571328269
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,256,0.13748710685306126
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,4096,64,0.0027946666296985415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,4096,128,0.011239111423492432
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,128,0.13494577672746447
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,4096,32,0.0026480000880029467
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,4096,64,0.011585777832402123
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3584,65536,0.07932621902889676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,4096,32,0.010696888797812991
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3584,65536,0.0678222245640225
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3584,16384,0.0243520008193122
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,65536,0.06656177838643391
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3584,16384,0.02537422213289473
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,65536,0.1500062280231052
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3584,12288,0.01904177831278907
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,16384,0.020216888851589628
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3584,12288,0.025103110406133864
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,512,0.005902222047249476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,12288,0.140557328859965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3584,10240,0.017143110434214275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,12288,0.016542222764756944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,4096,3072,0.00998577806684706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3584,8192,0.014921777778201632
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3584,10240,0.02277066641383701
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,10240,0.015073777900801765
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,4096,128,0.005720888988839255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3584,7168,0.013432888521088494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3584,8192,0.021943999661339655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,8192,0.14016711711883545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3584,7168,0.021018667353524104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3584,6144,0.012703999876976013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,7168,0.011374221907721626
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,7168,0.1409671174155341
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3584,6144,0.01977599991692437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,6144,0.010728889041476779
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3584,5120,0.011373333632946014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,16384,0.14381244447496203
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3584,5120,0.018810666269726224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,5120,0.14068267080518934
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3584,4096,0.01036088913679123
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,5120,0.009689778089523315
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3584,4096,0.017521777086787753
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3584,3584,0.011335999601417117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,4096,0.14069155851999918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3584,3584,0.01583377851380242
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,3584,0.1399537722269694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,10240,0.14482577641805014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3584,3072,0.0099982221921285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,3584,0.008051555189821456
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,8192,0.01278222186697854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3584,3072,0.014121777481502958
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,3072,0.1397093269560072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,3072,0.007658667034573025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3584,2560,0.01348088930050532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,2560,0.007299555672539606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,6144,0.1391484472486708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3584,2048,0.013735111388895245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,2048,0.138081775771247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,2048,0.006966222491529252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3584,1536,0.005937777873542573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3584,1536,0.013052444491121503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,1536,0.13722933663262263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,4096,0.00831822223133511
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3584,1024,0.004917333523432414
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,1536,0.006552888701359431
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3584,1024,0.012427555190192329
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,4096,8192,0.14555554919772679
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,1024,0.1378177801767985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,1024,0.006241777704821692
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3584,768,0.011716444459226398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3584,2560,0.008388444781303406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,768,0.13923554950290257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3584,512,0.003867555409669876
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,512,0.005952888892756567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3584,256,0.0028275555620590844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3584,2048,0.007265778051482306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3584,256,0.010438222024175854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,256,0.00590844452381134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,256,0.14167200194464788
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3584,128,0.0028355556229750314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3584,128,0.011029333704047732
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,128,0.005642666584915585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3584,64,0.0028159999185138275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3584,32,0.002534222271707323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3584,64,0.011045333411958484
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3584,32,0.011031111081441244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3072,65536,0.06310222546259563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3584,768,0.00423555572827657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3072,65536,0.0650177796681722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3072,16384,0.02051377793153127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,3584,768,0.006217777729034424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,65536,0.06529600090450711
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,65536,0.14942044681972927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3072,16384,0.02443644404411316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3072,12288,0.016931555337376065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3584,512,0.011745777395036487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,16384,0.1401288906733195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,2560,0.13933332761128744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,16384,0.019928889142142404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3072,10240,0.015488000379668342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3072,12288,0.024375110864639282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,12288,0.01648000048266517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,12288,0.1469928953382704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3072,8192,0.012866666747464074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3072,10240,0.022669333550665114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,10240,0.14105155732896593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,10240,0.01441422270403968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3072,7168,0.011852444046073489
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3072,8192,0.020592888196309406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,8192,0.01241422196229299
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,8192,0.14040266142951116
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3072,6144,0.011200889117187925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3072,7168,0.019334221879641216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,7168,0.011067555182509951
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,7168,0.14027644528283015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3072,5120,0.010866666833559671
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3072,6144,0.018226666582955253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,6144,0.010050666828950247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,6144,0.1461928950415717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3072,4096,0.00962933318482505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3072,5120,0.016568889220555622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,5120,0.14587377177344427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3072,4096,0.014896889527638754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,4096,0.008291555775536431
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3072,3584,0.014564444621404013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,3584,0.13951999611324734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3072,3072,0.008933333059151968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,512,0.13801688618130153
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3072,3072,0.014106666048367819
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,3072,0.13888622654808894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3072,2560,0.007749333149856991
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,3072,0.007606222397751278
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,3584,128,0.13419022825029162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3072,2560,0.01346933344999949
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3072,2048,0.006863111009200414
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,2560,0.007263999846246507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3072,2048,0.012881778180599213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,2048,0.14042488733927408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,2048,0.007032000356250339
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3072,1536,0.006216889040337668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3072,3584,0.00870488915178511
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,4096,0.14534933037228054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3072,1024,0.004863111095296012
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3072,1536,0.012703999876976013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,1536,0.13777155346340603
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3072,1024,0.012050666742854647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,1024,0.006224000205596288
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,3584,0.007967111137178209
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,1024,0.14043111271328396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3072,768,0.00452533322903845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3072,512,0.003846222327815162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,768,0.006067555397748947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,768,0.14444888962639704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3072,512,0.011501332951916588
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3072,256,0.00349866681628757
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,512,0.13913155926598444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3072,256,0.011367999845080905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,2560,0.13979732990264893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,256,0.13869688245985243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3072,128,0.002879111096262932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,256,0.0059075554211934405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3072,128,0.011052444577217102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,128,0.005606222069925732
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,3072,128,0.13280533419715032
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,5120,0.009471110999584198
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3072,32,0.0027955555253558685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2560,65536,0.05708355373806424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3072,64,0.011701333026091257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3072,32,0.011360000405046674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,1536,0.006603555546866522
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2560,16384,0.019010666343900893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2560,65536,0.0628604425324334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,65536,0.06439822249942355
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2560,16384,0.023904000719388325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2560,12288,0.015438222222858004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,3072,768,0.012083555261294046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,16384,0.13977600468529597
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,16384,0.01958044370015462
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2560,10240,0.013950222068362765
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2560,12288,0.023256889647907678
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,12288,0.01612977849112617
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,12288,0.14164888858795166
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2560,8192,0.011748444702890186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2560,10240,0.022608000371191237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,10240,0.14706399705674914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,10240,0.014138667119873894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2560,7168,0.011376889215575324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2560,8192,0.01977244516213735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,8192,0.1397884421878391
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,8192,0.011725333001878528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2560,6144,0.010636444720957013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2560,7168,0.018969777557584975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,7168,0.011357333097192975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,7168,0.14088533984290227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2560,5120,0.009712888962692684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2560,6144,0.01717422240310245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,6144,0.009979555176364051
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,6144,0.13962754938337538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2560,5120,0.016509332590632968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2560,4096,0.009076444639099969
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,5120,0.009339555270142024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2560,4096,0.014663999279340109
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2560,3584,0.008612444831265343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,4096,0.00830666638082928
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,4096,0.14098132981194392
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,65536,0.1487857765621609
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2560,3072,0.008267555799749162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2560,3584,0.014487110906177096
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,3584,0.008005333443482717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,3584,0.138629330529107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2560,2560,0.0069057775868309875
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,3072,512,0.005974222388532426
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2560,3072,0.013458666702111563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,3072,0.14130844010247126
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2560,2048,0.006222222414281633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2560,2560,0.013447110851605734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,2560,0.0073315559162033936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,2560,0.14508267243703207
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2560,1536,0.005593777944644292
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2560,2048,0.013342222405804528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,2048,0.0069208890199661255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,2048,0.13997333579593235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2560,1024,0.004849777867396672
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,3072,64,0.0028000000036425064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2560,1536,0.012771555946932899
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,1536,0.006640000061856375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,1536,0.14309155941009521
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2560,768,0.004184000194072723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2560,1024,0.012373333175977072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,1024,0.13804622491200766
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2560,512,0.0038559999730851916
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2560,768,0.012096888489193387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,768,0.005985777825117111
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,5120,0.14054132832421198
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2560,256,0.0031644445326593188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2560,512,0.011394666300879585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,512,0.1426924467086792
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2560,256,0.011122666299343109
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,256,0.0058906666106647914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2560,128,0.003137777869900068
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2560,128,0.011054221954610614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,128,0.1330826679865519
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,3072,0.007601777712504069
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2560,64,0.0027911110470692315
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2560,32,0.0028000000036425064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,128,0.005587555468082428
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2560,64,0.011723555624485016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2048,65536,0.050075554185443454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2560,32,0.011035555766688453
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2048,65536,0.06033155653211805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2048,16384,0.015695111619101632
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,65536,0.059213333659701876
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,1024,0.00629333323902554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,65536,0.14982222186194524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2048,12288,0.013037333057986366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,16384,0.140883551703559
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,768,0.14383110735151503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,16384,0.01954844428433312
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2048,10240,0.011736888852384357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2048,12288,0.022142221530278523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,2560,512,0.005919111271699269
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,12288,0.014427555931939019
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2048,8192,0.011356444822417365
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2048,10240,0.02055022286044227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,2560,256,0.14100978109571669
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,10240,0.14308533403608534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2048,8192,0.0184568895234002
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,8192,0.011405333048767514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,8192,0.14496533075968424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2048,6144,0.009680888719028896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2048,7168,0.017900443739361234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,7168,0.1446337832344903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,7168,0.0107004443804423
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2048,5120,0.009093333449628618
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2048,6144,0.015812445018026564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,6144,0.14234933588239881
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,6144,0.01000088867213991
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2048,4096,0.010252444280518426
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2048,5120,0.0151582227812873
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,5120,0.009301332963837517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,5120,0.14396444956461588
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2048,3584,0.009576000273227692
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2048,4096,0.014104889498816596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,4096,0.0083137775460879
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,4096,0.14568889141082764
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2048,3072,0.006621333460013072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2048,3584,0.013752889302041797
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,3584,0.1467697752846612
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,3584,0.007992000215583378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2048,2560,0.0058959999846087555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2048,3072,0.013776889277829064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,3072,0.14420444435543484
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,12288,0.14055200417836508
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2048,2560,0.013111111190583972
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2048,2048,0.0069066666894488865
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,2560,0.1440702279408773
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,2560,0.0074746666683091064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,10240,0.013975999421543546
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2048,1536,0.0058497778243488735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2048,2048,0.013074666261672974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,2048,0.006973333656787872
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,2048,0.1382897827360365
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2048,1024,0.004583111239804162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2048,1536,0.012735111018021902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,1536,0.006631999793979857
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,1536,0.14229955938127306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2048,768,0.004182222402758068
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2048,1024,0.012372444073359171
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,1024,0.006235555642180973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,1024,0.1387502219941881
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2048,512,0.003539555602603488
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2048,768,0.012039999994966718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,768,0.13917511039310032
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,768,0.0059075554211934405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2048,256,0.0032142222755485107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2048,512,0.011704000333944956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,512,0.0058764442801475525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,512,0.1414711078008016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2048,128,0.00286311118139161
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2048,256,0.011266666981909009
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,256,0.005559999909665849
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,256,0.1387759976916843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2048,64,0.0028479999552170434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2048,128,0.011063999599880643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2048,32,0.002814222127199173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,2048,128,0.13677333460913763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,3072,0.007652444144090016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2048,64,0.011681777735551199
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1536,65536,0.03597777750757005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2048,32,0.011090666883521609
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1536,16384,0.013515555196338229
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1536,65536,0.058268447717030845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,2048,7168,0.010649777948856354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,65536,0.0603804455863105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1536,16384,0.023008000519540574
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,65536,0.148272885216607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,16384,0.01846933364868164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,16384,0.1468755536609226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1536,10240,0.010020444790522257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1536,12288,0.02163822286658817
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,12288,0.01447111037042406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,12288,0.13965511322021484
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1536,8192,0.0103493332862854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1536,10240,0.01927911076280806
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,10240,0.14245866404639349
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,10240,0.013125333521101209
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1536,8192,0.016558221644825406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,8192,0.14637422561645508
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,8192,0.01130222198035982
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1536,7168,0.015882665912310284
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1536,6144,0.008968888885445064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,2048,16384,0.023665777511066858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,7168,0.14497955640157065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,2048,128,0.005602666487296422
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1536,6144,0.015451555450757345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,6144,0.009783999787436591
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,6144,0.14151022169325086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1536,5120,0.014783110883500842
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,5120,0.009025777379671732
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1536,4096,0.007288888924651676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1536,12288,0.011534222298198275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1536,4096,0.013800000150998434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,4096,0.1418773333231608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,4096,0.008292444050312042
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1536,3584,0.013437333206335703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,3584,0.1449715561336941
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,3584,0.007976888782448238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1536,3072,0.006234666539563074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1536,7168,0.009448889229032729
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1536,3072,0.013441777891582914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,3072,0.007704888780911763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1536,2560,0.005912888795137405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,7168,0.01034577770365609
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1536,2560,0.013070222404268054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,2560,0.14014577865600586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1536,5120,0.008375110725561777
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,2560,0.007264888948864407
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1536,2048,0.012715555727481842
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,2048,0.0069617778062820435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1536,1536,0.005202666752868229
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,5120,0.1471626626120673
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1536,1536,0.012727999852763282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1536,1024,0.004219555606444676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,1536,0.006599999964237213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,1536,0.13881956206427679
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1536,3584,0.007280000381999546
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1536,768,0.0038604442444112566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1536,1024,0.012383999923865
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,1024,0.14624533388349745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1536,768,0.011887999872366587
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1536,512,0.00351555562681622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,768,0.14131377802954778
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1536,512,0.011377777490350934
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,512,0.14199911223517522
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1536,256,0.0031786666562159858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,3072,0.14497244358062744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,512,0.0058959999846087555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1536,256,0.011185777684052786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1536,128,0.002826666666401757
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1536,2048,0.006220444622966979
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1536,128,0.011367110742463006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,128,0.13087911076015896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,2048,0.13864978154500326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1536,64,0.0024951110697454875
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1536,32,0.0025128889828920364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,128,0.005607111172543631
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1024,65536,0.02528800070285797
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1536,64,0.01202400028705597
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1536,32,0.0107342220014996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,1024,0.006274666637182236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1024,65536,0.05483288897408379
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1024,16384,0.014120888378885059
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,65536,0.1479111115137736
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,768,0.005935110979610019
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,16384,0.016857778032620747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1024,12288,0.012807110945383707
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1024,12288,0.01909244391653273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,12288,0.1463466617796156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1024,10240,0.011429333024554782
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,12288,0.01382844481203291
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,1536,256,0.005630222045713001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1024,10240,0.017824888229370117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1024,8192,0.010198222266303169
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,10240,0.012667555775907306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1024,8192,0.01629066632853614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,8192,0.14154222276475695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1024,7168,0.009887999958462184
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,8192,0.011039111349317761
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1024,7168,0.015803555647532146
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,7168,0.1460826661851671
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,65536,0.058732443385654025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1024,6144,0.008656889200210571
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,7168,0.01036800030204985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1024,16384,0.022641777992248535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,16384,0.13884088728162977
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1024,5120,0.008270222279760573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1024,6144,0.015474667151769003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,6144,0.009992889232105678
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1024,4096,0.007273777491516537
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,5120,0.009333333207501305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1024,5120,0.014766222900814481
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,5120,0.14289422829945883
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1024,3584,0.006911110960774952
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1024,4096,0.013758222262064615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,4096,0.00831822223133511
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,4096,0.13952355914645725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,1536,256,0.14389422204759386
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1024,3072,0.006224000205596288
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1024,3584,0.013791110780504016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,3584,0.00793688909875022
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,3584,0.1410586701499091
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1024,2560,0.005631111148330901
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,10240,0.14121333758036295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1024,3072,0.013471110827393003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,3072,0.007596444752481248
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1024,2048,0.006252444452709622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1024,2560,0.013091555900043912
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,2560,0.13954844739702013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1024,2048,0.012870222330093384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1024,1536,0.005544888890451855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,2048,0.14476354916890463
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,2048,0.00692622239391009
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1024,1024,0.00452888881166776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1024,1536,0.01275288893116845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,1536,0.006636444479227066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1024,1024,0.012092444631788464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,1024,0.006263111200597551
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,1024,0.1412924395667182
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,6144,0.14625867207845053
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1024,768,0.012067555553383298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,768,0.14236356152428523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1024,512,0.00351555562681622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,768,0.0059279998143513995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1024,512,0.011739555332395764
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1024,256,0.0031511110977994073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,512,0.14501510726081
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,512,0.005912888795137405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1024,256,0.011357333097192975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,256,0.005583999885453119
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,256,0.14432533582051596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,3072,0.1382675568262736
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,2560,0.007294221884674496
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1024,128,0.01035822182893753
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1024,64,0.002879111096262932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,128,0.13488266203138563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,1024,128,0.005611555443869696
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1024,32,0.0028000000036425064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,768,65536,0.021402666966120403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1024,64,0.01075111081202825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,1024,32,0.010752889017264048
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,768,16384,0.01034044474363327
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,1024,1536,0.13939910464816624
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,768,65536,0.05511022276348538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,65536,0.14676445060306126
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,65536,0.05816000037723117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,768,12288,0.010808889236715106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,768,16384,0.021371554997232225
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,16384,0.016864000095261466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,16384,0.14584622118208143
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,768,10240,0.010019555687904358
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,768,12288,0.01886133352915446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,12288,0.14529955387115479
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,12288,0.014088888963063559
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,768,8192,0.008982222113344405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,768,10240,0.017509332961506315
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,10240,0.012710221939616732
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,10240,0.13976977931128606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,768,7168,0.008632000121805403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,768,8192,0.016166221764352586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,8192,0.011047110789351992
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,768,6144,0.00796888851457172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,768,7168,0.015807999504937064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,7168,0.01038577738735411
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,7168,0.14006933901045057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,768,5120,0.007078222102589077
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,768,6144,0.014947555131382413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,6144,0.13937066660987005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,6144,0.009753777749008602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,768,4096,0.0074924445814556545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,768,5120,0.0148026661740409
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,5120,0.009283555878533257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,768,4096,0.013820444544156393
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,768,3584,0.0069448889957533936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,4096,0.1414817836549547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,4096,0.00828888929552502
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,3584,0.007910221815109253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,3584,0.14080888695187038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,768,3072,0.006601777755551868
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1024,768,0.0042071110672420925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,768,3072,0.013750221994188098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,3072,0.14056089189317492
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,768,2560,0.006270222365856171
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,768,2560,0.013049778011110095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,2560,0.14094666639963785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,768,2048,0.005195555587609609
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,2560,0.007263110743628607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,768,2048,0.012753778033786349
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,1024,128,0.0028213332924577924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,2048,0.13941510518391928
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,768,1536,0.004912888838185204
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,2048,0.0069208890199661255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,5120,0.1459991137186686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,768,1024,0.004209777961174647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,1536,0.14203288820054796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,768,1024,0.011749332977665795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,768,3584,0.013419555293189155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,768,768,0.0038417776425679526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,1024,0.0059262220230367445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,768,768,0.011739555332395764
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,3072,0.007607999775144789
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,768,0.14504178365071616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,768,512,0.003519111209445529
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,768,0.005952000204059813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,768,512,0.011061333119869232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,8192,0.14467467202080622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,768,256,0.0028524444335036804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,512,0.14038666089375815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,512,0.005853333406978183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,768,128,0.0028088889602157804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,768,256,0.011393778026103973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,256,0.005609777652555042
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,768,128,0.011012444065676795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,768,64,0.0026373333401150177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,128,0.13656089040968153
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,128,0.005590222362014983
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,768,32,0.002504000026318762
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,512,65536,0.01701866587003072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,768,1536,0.012431110772821637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,768,64,0.011713777979214987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,768,1536,0.006559111177921295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,768,32,0.011032889286677042
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,512,16384,0.011039999624093374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,512,65536,0.05361422234111362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,65536,0.14736444420284694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,65536,0.05608355336719089
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,1024,0.1425173282623291
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,512,16384,0.02093688812520769
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,512,12288,0.00871733327706655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,16384,0.14071644677056208
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,512,10240,0.008063111040327284
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,16384,0.01682577696111467
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,512,12288,0.018681777848137748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,12288,0.013789333403110504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,512,8192,0.007986666427718269
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,512,10240,0.01771111124091678
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,10240,0.01276088919904497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,512,7168,0.007663110891977946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,512,8192,0.016196444630622864
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,8192,0.011187555889288584
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,512,7168,0.01588711142539978
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,512,6144,0.007285333342022366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,7168,0.14185778299967447
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,768,256,0.14444355169932047
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,512,6144,0.015206222732861837
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,6144,0.1402346690495809
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,512,5120,0.006948444578382704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,6144,0.009680888719028896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,512,5120,0.014840000205569796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,5120,0.14129689004686144
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,5120,0.00903733323017756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,512,4096,0.013773333695199756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,4096,0.1459804508421156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,512,3584,0.0069306666652361555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,4096,0.00833333366447025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,12288,0.14306399557325575
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,512,3584,0.013440888788965015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,10240,0.14062044355604383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,3584,0.14144710699717203
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,512,3072,0.006613333192136552
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,3584,0.007989333735571967
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,8192,0.14696621894836426
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,512,2560,0.006268444574541516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,512,3072,0.013448889056841532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,3072,0.007632888853549957
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,3072,0.1413795550664266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,512,2048,0.005921777751710679
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,7168,0.01035911093155543
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,2560,0.007303111255168915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,512,2560,0.013066666821638743
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,2560,0.14226222038269043
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,512,1536,0.004509333521127701
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,512,2048,0.012441777520709567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,2048,0.14783199628194174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,512,1536,0.012387555506494312
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,1536,0.1419244474834866
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,512,1024,0.011126221881972419
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,512,768,0.003528000166018804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,512,4096,0.007296889192528195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,1024,0.005981333139869902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,1024,0.14478488763173422
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,512,512,0.0032106666929192017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,768,0.0061048890153567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,768,0.14046755101945665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,512,512,0.011393778026103973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,512,256,0.0031679999083280563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,512,0.005889777921968036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,512,0.14396622445848253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,512,256,0.011024889018800525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,512,128,0.002806222273243798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,256,0.005587555468082428
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,256,0.13836799727545843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,512,128,0.011027555498811932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,128,0.005576888720194499
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,512,128,0.1332471105787489
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,512,64,0.0028080000645584534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,512,32,0.0028497777465316984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,256,65536,0.015648888217078317
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,512,64,0.011400000088744693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,512,32,0.01089599976936976
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,2048,0.0069306666652361555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,256,16384,0.009301332963837517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,65536,0.14697599411010742
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,256,16384,0.020968000094095867
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,16384,0.13968443870544434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,16384,0.01685333251953125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,512,768,0.01108088923825158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,256,12288,0.018806222412321303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,12288,0.14698844485812718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,12288,0.013943110903104147
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,256,10240,0.017196445001496207
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,10240,0.1464115513695611
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,256,8192,0.0069546666410234236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,10240,0.012632889052232107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,256,8192,0.016529776983790927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,8192,0.1481635570526123
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,512,1024,0.0038382220599386427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,256,7168,0.006596444381607904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,256,65536,0.051318221622043185
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,8192,0.011114666859308878
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,256,6144,0.006609777609507243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,65536,0.050880889097849526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,256,7168,0.015797333584891427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,7168,0.010317333042621613
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,7168,0.14596088727315268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,256,5120,0.006260444306664997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,256,6144,0.01514755520555708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,6144,0.009849777652157677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,256,4096,0.005904888941182031
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,256,12288,0.008285333712895712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,256,5120,0.014447111222479077
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,5120,0.14191910955641004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,5120,0.009359111388524374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,256,3584,0.006296000132958095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,4096,0.008001777860853408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,256,4096,0.014127110441525778
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,4096,0.1417964431974623
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,256,10240,0.007615110940403408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,256,3584,0.013788444300492605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,3584,0.14103733168707952
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,3584,0.00775822252035141
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,256,2560,0.006270222365856171
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,256,3072,0.01328000012371275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,3072,0.007631999750932057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,3072,0.14092532793680826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,256,2048,0.005938666562239329
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,256,2560,0.01238933371173011
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,2560,0.007285333342022366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,2560,0.1420960028966268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,256,1536,0.004580444345871608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,2048,0.14509956041971842
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,2048,0.006953777538405524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,256,1536,0.01276533305644989
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,256,1024,0.003891555385457145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,1536,0.006616000086069107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,512,1536,0.006625777731339137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,1536,0.14493600527445474
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,256,768,0.0035217776894569397
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,1024,0.006247111078765657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,1024,0.14016711711883545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,256,768,0.012083555261294046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,6144,0.14557422531975642
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,256,512,0.0031679999083280563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,768,0.13973955313364664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,256,512,0.011356444822417365
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,512,0.0058702222175068325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,512,0.14512622356414795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,256,3072,0.005919999960396025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,256,256,0.011104889214038849
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,256,0.14435022407107884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,256,128,0.0028115556471877625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,256,0.005558222118351195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,128,0.00565155554148886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,256,128,0.011092444260915121
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,256,128,0.13192000654008654
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,256,32,0.002526222210791376
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,256,64,0.012087999946541257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,256,32,0.011054221954610614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,128,65536,0.04930044545067681
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,256,1024,0.012093333734406365
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,65536,0.14515554904937744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,256,768,0.006076444354322221
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,128,16384,0.007668444679843054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,256,256,0.0028764444092909494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,128,16384,0.02093066606256697
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,16384,0.13935200373331705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,16384,0.016747555798954435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,128,12288,0.018519111805491976
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,12288,0.14619466993543836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,12288,0.013781333135233985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,128,10240,0.007261333366235097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,256,2048,0.01276533305644989
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,128,10240,0.01753777762254079
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,10240,0.012696888711717395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,128,8192,0.006597333484225803
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,128,8192,0.01647644407219357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,8192,0.13936532868279353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,128,7168,0.0063786668082078295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,8192,0.01108000013563368
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,65536,0.05019466744528877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,128,7168,0.015626667274369132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,128,6144,0.006333333336644703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,7168,0.14228622118631998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,128,12288,0.007428444094128079
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,128,6144,0.014893333117167154
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,128,5120,0.005888000130653381
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,6144,0.13960177368587917
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,6144,0.009710222482681274
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,128,5120,0.01442577772670322
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,5120,0.13920177353752985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,256,64,0.0025128889828920364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,128,4096,0.014067555467287699
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,128,65536,0.014431110686726041
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,128,3584,0.006280000011126201
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,4096,0.008279110822412703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,128,3584,0.013787555197874704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,3584,0.1389626661936442
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,3584,0.00796977761718962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,128,3072,0.013457777599493662
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,3072,0.14544533358679876
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,7168,0.0103493332862854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,128,2560,0.005604444278611078
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,3072,0.007647111184067196
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,128,2560,0.01311377767059538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,2560,0.007303111255168915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,128,2048,0.012816888590653738
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,2048,0.13861600557963052
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,10240,0.1464808914396498
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,128,1536,0.0042364444169733255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,2048,0.006949333681000604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,128,1536,0.012435555458068848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,128,3072,0.005949333310127258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,128,1024,0.00387999994887246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,1536,0.006631999793979857
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,128,1024,0.012091555529170565
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,1024,0.14466933409372965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,128,768,0.0034097778714365433
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,1024,0.005993777679072485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,2560,0.14646044042375353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,128,768,0.011763555308183035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,128,512,0.0031600000543726813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,768,0.14419821898142496
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,768,0.005920889063013925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,5120,0.009311111436949836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,128,256,0.0028239999794297745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,128,512,0.01143022212717268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,512,0.005557333429654439
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,512,0.1449848943286472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,128,256,0.011187555889288584
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,4096,0.13991199599372015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,256,0.14403377638922796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,128,128,0.002514666567246119
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,256,0.005575110928879843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,128,64,0.0025111111915773815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,8,128,128,0.005620444400442972
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,128,128,0.011033777561452655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,128,0.14084088802337646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,128,64,0.012084444363911947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,64,65536,0.010696888797812991
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,8,128,1536,0.13887200090620253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,64,16384,0.007270221908887227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,128,32,0.010701333483060202
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,64,65536,0.049265778727001615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,64,16384,0.020976000361972388
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,64,10240,0.006335111127959357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,64,12288,0.018578666779730055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,64,8192,0.006616000086069107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,64,7168,0.006331555545330048
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,64,10240,0.017480888300471835
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,64,6144,0.005930666708283954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,128,2048,0.005922666854328579
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,64,7168,0.015825778245925903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,64,5120,0.005588444570700328
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,64,4096,0.005581333405441708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,64,6144,0.015438222222858004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,128,4096,0.005859555469618903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,64,3584,0.006215111249023014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,64,3072,0.005931555396980709
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,64,4096,0.014124444789356656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,64,2560,0.00628355559375551
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,64,3584,0.013470222552617392
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,64,2048,0.005869333528810077
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,64,3072,0.01347733371787601
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,64,1536,0.004231111043029361
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,64,2560,0.013064000341627332
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,64,1024,0.0038702223036024305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,64,2048,0.012727110750145383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,64,768,0.003488000068399641
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,128,32,0.002536888958679305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,64,1536,0.012444444828563266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,64,512,0.003169777699642711
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,64,1024,0.012111110819710625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,64,256,0.0028737777223189673
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,64,128,0.002504000026318762
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,64,12288,0.006272888845867581
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,64,768,0.01203644441233741
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,64,512,0.011688888900809817
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,64,64,0.002461333241727617
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,64,32,0.002508444504605399
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,64,256,0.011076444553004371
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,32,65536,0.009335111412737105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,64,128,0.011047110789351992
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,64,64,0.011392888923486074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,32,16384,0.006232000059551663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,64,32,0.010851555400424533
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,32,12288,0.006217777729034424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,32,65536,0.0493182208802965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,32,10240,0.005947555518812603
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,32,8192,0.005595555735958948
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,64,8192,0.016203555795881484
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,32,7168,0.0059262220230367445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,32,6144,0.00592533333433999
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,32,8192,0.016148444679048326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,32,5120,0.0058782220714622075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,32,7168,0.015561777684423657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,32,6144,0.015463110473420886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,32,4096,0.005519999812046687
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,32,3584,0.0058666666348775225
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,32,5120,0.014428445034556918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,32,3072,0.005882666756709416
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,32,4096,0.014143110977278816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,32,2560,0.005939555664857228
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,32,3584,0.013443555268976422
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,32,2048,0.005633777628342311
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,32,3072,0.013150222599506378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,32,1536,0.004536889079544279
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,32,2560,0.01313511116637124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,32,1024,0.003899555653333664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,32,768,0.003547555456558863
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,32,1536,0.01238044434123569
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,32,512,0.003133333391613431
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,32,768,0.012238221863905588
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,32,256,0.002862222285734283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,32,512,0.011360000405046674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,32,128,0.002481777841846148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,32,256,0.011364444262451597
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,32,64,0.00258577780591117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,32,16384,0.02092711130777995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,32,128,0.011019555230935415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,32,12288,0.018698667486508686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,32,10240,0.017450667089886136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,32,64,0.012039110892348819
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,32,32,0.010692444112565784
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,64,5120,0.014532445205582513
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,65536,16384,0.338442669974433
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,65536,16384,0.24112354384528267
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,65536,12288,0.259530676735772
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,32,2048,0.012692444854312472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,16384,0.13719911045498318
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,65536,12288,0.19459022416008842
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,8,32,1024,0.011071999867757162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,65536,10240,0.21727555327945283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,65536,10240,0.1778115563922458
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,12288,0.10436266660690308
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,8,32,32,0.002487999904486868
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,65536,8192,0.16836088233523896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,65536,8192,0.12650222248501247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,10240,0.08791555298699273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,65536,7168,0.14759288893805608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,12288,0.16975555155012342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,8192,0.06960088676876493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,65536,7168,0.11260888311598037
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,16384,0.213137772348192
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,10240,0.15101777182685003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,65536,6144,0.12819822629292807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,65536,6144,0.09989600049124824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,7168,0.06326755550172594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,65536,5120,0.1033066643608941
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,8192,0.14236533641815186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,65536,4096,0.085789336098565
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,6144,0.0735786689652337
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,5120,0.046888887882232666
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,65536,4096,0.07124800152248807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,6144,0.1427075597974989
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,65536,3584,0.07276622454325359
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,5120,0.13999822404649523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,4096,0.03863288958867391
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,65536,3584,0.06588177548514472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,65536,3072,0.06357066498862372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,4096,0.14609511693318686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,3584,0.034875555170906916
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,65536,3072,0.06138222085105049
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,3072,0.03117155697610643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,3584,0.1408968898985121
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,65536,2560,0.05323822299639384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,3072,0.13967822657691106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,2560,0.02846577763557434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,2560,0.1469120052125719
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,65536,5120,0.08768888976838853
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,65536,2048,0.039804445372687444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,65536,1536,0.03327200147840712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,2048,0.027973333994547527
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,7168,0.1415519979265001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,2048,0.14098578029208714
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,65536,1536,0.03591377867592706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,65536,1024,0.024002666274706524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,1536,0.023678221636348303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,1536,0.14402488867441812
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,65536,768,0.018781332506073844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,65536,1024,0.023611555496851604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,1024,0.018210666047202218
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,1024,0.14086400138007268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,65536,512,0.013758222262064615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,65536,768,0.0237128883600235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,768,0.015146666102939181
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,65536,256,0.008598222500748104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,65536,512,0.01924711134698656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,512,0.1391564475165473
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,65536,2560,0.05358933409055074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,65536,128,0.006532444308201472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,512,0.012997333374288348
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,65536,256,0.01608711149957445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,256,0.14057421684265137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,256,0.010448888772063784
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,65536,64,0.0058168888919883305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,65536,32,0.005613333235184352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,65536,128,0.014787554740905762
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,128,0.13529955016242132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,65536,128,0.01051644484202067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,65536,2048,0.04381422201792399
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,65536,32,0.014243554737832812
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,65536,64,0.014787554740905762
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,16384,16384,0.08626933230294122
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,16384,65536,0.18161333931816948
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,16384,16384,0.07493599918153551
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,16384,65536,0.3228071000840929
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,16384,12288,0.06687377558814155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,16384,0.04463466670778063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,65536,0.16403733359442815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,16384,12288,0.05830844243367513
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,16384,10240,0.05660177601708306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,12288,0.03465155429310269
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,65536,768,0.14511111047532824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,16384,0.1400231122970581
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,16384,8192,0.043792890177832715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,16384,10240,0.0558168888092041
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,12288,0.14135644170973036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,10240,0.0303457776705424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,16384,7168,0.03692355420854356
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,8192,0.024877332978778418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,16384,8192,0.04347733325428433
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,16384,6144,0.032919999625947736
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,8192,0.13956711027357313
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,7168,0.02219644520017836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,16384,6144,0.038562665383021034
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,16384,5120,0.027837332752015855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,6144,0.019876443677478366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,6144,0.14580267005496556
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,16384,5120,0.032448000378078885
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,16384,4096,0.023189332750108507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,5120,0.14162400033738878
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,5120,0.022014222211307947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,65536,0.21697955661349824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,16384,4096,0.02714311083157857
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,4096,0.015809777710172866
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,4096,0.14009777704874674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,16384,3584,0.020625778370433383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,16384,3072,0.01790577835506863
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,3584,0.014389332797792224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,16384,3584,0.025459556115998164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,3584,0.14451200432247585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,16384,3072,0.02358666724628872
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,16384,2560,0.015804444750150044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,3072,0.013194666968451606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,3072,0.13905599382188585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,16384,2048,0.01273155543539259
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,16384,2560,0.022328888376553852
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,16384,7168,0.04207288887765673
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,2560,0.14380978213416204
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,2560,0.012719111310111152
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,16384,1536,0.010382222632567087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,16384,2048,0.01957155598534478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,7168,0.1449528932571411
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,2048,0.1410542196697659
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,16384,1536,0.018043556147151522
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,1536,0.14377955595652261
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,16384,1024,0.014885332849290637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,16384,768,0.007019555403126611
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,1024,0.14530311690436468
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,1024,0.009119110802809397
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,16384,768,0.01461955573823717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,768,0.14493689272138807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,16384,512,0.005557333429654439
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,768,0.00832177781396442
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,512,0.007296000089910295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,16384,512,0.012239110966523489
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,512,0.14115199777815077
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,16384,256,0.003478222423129611
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,16384,256,0.011383111278216044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,256,0.006640888750553131
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,16384,128,0.0029333333174387612
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,16384,128,0.012426666915416718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,10240,0.1411679983139038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,128,0.13918044832017687
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,16384,64,0.0028231110837724474
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,16384,32,0.002874666617976295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,128,0.006781333436568578
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,2048,0.01238666640387641
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,16384,1024,0.008105778031879002
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,16384,1536,0.010958222051461538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,16384,32,0.010718222293588849
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,12288,16384,0.06644533077875774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,12288,65536,0.1584871080186632
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,12288,16384,0.0648533304532369
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,65536,0.14180178112453884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,12288,12288,0.051502221160464816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,16384,0.03943822118971083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,16384,256,0.13957066006130642
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,12288,12288,0.05236177643140157
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,12288,10240,0.04407377706633674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,12288,0.030810667408837214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,12288,0.1411679983139038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,12288,10240,0.045574221346113414
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,12288,8192,0.03881599836879306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,16384,64,0.011045333411958484
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,10240,0.02658933401107788
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,65536,0.1779902246263292
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,10240,0.14070755905575222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,12288,8192,0.0399582220448388
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,12288,7168,0.03472888800832961
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,8192,0.022302221920755174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,8192,0.147361781862047
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,12288,6144,0.030429333448410034
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,12288,7168,0.038089778688218855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,7168,0.1398799949222141
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,12288,6144,0.034576889541414045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,12288,5120,0.026918222506841023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,6144,0.018361777067184448
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,6144,0.14023289415571424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,12288,4096,0.023004444109068975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,12288,5120,0.029655112160576716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,16384,0.1440017753177219
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,5120,0.13991910881466335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,12288,3584,0.020483555065260995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,12288,4096,0.02609777781698439
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,4096,0.014429334137174817
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,4096,0.14467022154066297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,12288,3584,0.02418400016095903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,12288,3072,0.017879999346203275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,3584,0.14201422532399496
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,3584,0.013279111021094851
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,12288,2560,0.01566488875283135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,12288,3072,0.021976888179779053
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,3072,0.012145777543385824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,3072,0.1448648903104994
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,12288,2560,0.022037333912319605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,12288,2048,0.013792888985739814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,2560,0.011526222030321756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,2560,0.13982844352722168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,12288,65536,0.2421813276078966
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,12288,1536,0.008632889224423302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,2048,0.01106844428512785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,1536,0.009677333136399588
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,12288,1024,0.0070159998204973005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,12288,1024,0.01351999988158544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,5120,0.01647111111217075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,1024,0.008466666771305932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,1024,0.1398808823691474
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,12288,768,0.005589333259397083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,12288,768,0.014125333891974555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,768,0.1410577827029758
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,12288,512,0.003863111138343811
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,768,0.007623111208279927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,12288,512,0.013067555924256643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,512,0.13862844308217367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,12288,256,0.0031537777847713898
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,12288,2048,0.017881777551439073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,2048,0.1428648895687527
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,12288,256,0.012063999970753988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,1536,0.13881688647800022
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,12288,128,0.0027928888383838865
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,12288,1536,0.017584888471497428
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,128,0.006664888726340399
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,12288,128,0.011815999945004782
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,12288,64,0.0028373334142896864
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,12288,32,0.0028684443483750024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,12288,64,0.011350221931934357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,12288,32,0.010374222364690568
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,10240,65536,0.2043902211719089
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,10240,65536,0.14911111195882162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,512,0.007270221908887227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,10240,16384,0.05788177914089627
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,7168,0.020271999968422782
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,10240,16384,0.05840888950559828
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,65536,0.1375582218170166
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,128,0.1327528953552246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,10240,12288,0.0451528893576728
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,16384,0.037920888927247785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,16384,0.14346310827467176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,10240,12288,0.04558755622969734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,10240,10240,0.03905066516664293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,12288,0.02865955564710829
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,65536,0.16165333324008518
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,10240,8192,0.032405333386527166
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,10240,0.025404444999165002
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,10240,10240,0.04183822207980686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,10240,0.14018577999538845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,10240,7168,0.0292577776643965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,10240,8192,0.034456888834635414
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,8192,0.13915377193027073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,10240,7168,0.029631111356947158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,10240,6144,0.026345777842733595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,7168,0.1432924403084649
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,7168,0.019560888409614563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,10240,5120,0.02253066665596432
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,6144,0.13994044727749294
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,12288,256,0.006577777779764599
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,10240,5120,0.025361778007613286
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,10240,4096,0.019537778364287484
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,5120,0.01603200038274129
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,5120,0.13953867223527697
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,10240,4096,0.023223110371165808
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,4096,0.14279022481706408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,4096,0.013781333135233985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,10240,3584,0.02165511084927453
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,12288,0.14511111047532824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,3584,0.13789599471622044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,10240,3072,0.01626844373014238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,3584,0.013042666845851474
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,10240,3072,0.02090311050415039
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,8192,0.021372444099850122
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,10240,2560,0.0148062225845125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,3072,0.011778666741318174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,10240,2560,0.018976888722843595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,2560,0.010664888554149203
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,2560,0.13906221919589573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,10240,6144,0.028449777099821303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,10240,2048,0.010418666733635796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,10240,1536,0.008624888956546783
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,6144,0.017921777235137094
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,12288,256,0.14523200194040933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,10240,2048,0.01799199978510539
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,2048,0.010425777898894416
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,10240,1024,0.006236444330877728
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,2048,0.13931911521487766
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,1536,0.00906666699382994
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,10240,1536,0.015807110402319167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,1536,0.1422977712419298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,10240,768,0.005703111075692707
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,1024,0.14243200090196398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,1024,0.00794400026400884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,10240,512,0.0035413333939181436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,10240,768,0.013407111167907715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,768,0.007635555333561367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,10240,3584,0.017609778377744887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,10240,256,0.0031848889258172777
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,10240,512,0.012527999778588613
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,512,0.0069679998689227635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,512,0.1424915525648329
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,10240,128,0.00286666676402092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,256,0.14528621567620173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,256,0.00629333323902554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,10240,128,0.01236977759334776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,10240,64,0.0028453332682450614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,10240,128,0.006485333459244833
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,128,0.13618222872416177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,10240,32,0.0028355556229750314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,3072,0.145468446943495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,10240,64,0.011706666813956367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,10240,32,0.009732444253232744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,8192,65536,0.16939466529422334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,8192,65536,0.09806844260957505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,65536,0.07947022385067411
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,8192,16384,0.04599110947714912
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,10240,1024,0.013395555317401886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,8192,12288,0.033265779415766396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,16384,0.02379911144574483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,10240,768,0.14249689049190944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,8192,12288,0.036129779285854764
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,16384,0.14396622445848253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,8192,10240,0.028135110934575398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,10240,256,0.011680000358157687
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,12288,0.019141332970725167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,65536,0.15076533953348795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,12288,0.14616266886393228
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,8192,10240,0.03193688723776076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,10240,0.01719288859102461
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,8192,7168,0.021233777205149334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,8192,8192,0.02778844369782342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,8192,0.015076445208655464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,8192,0.14430756039089626
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,8192,6144,0.0188755558596717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,8192,7168,0.02533688810136583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,7168,0.01443822185198466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,7168,0.14399822552998862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,8192,5120,0.016122667325867545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,8192,6144,0.024459555745124817
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,8192,16384,0.04690133200751411
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,6144,0.1403368843926324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,6144,0.014178666803571912
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,8192,4096,0.0137155552705129
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,8192,5120,0.02238044473859999
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,5120,0.01203733351495531
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,5120,0.14701688289642334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,8192,3584,0.012294222083356647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,4096,0.1398577822579278
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,4096,0.01072355525361167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,8192,3072,0.011226666470368704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,8192,3584,0.019282667173279654
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,3584,0.1395688851674398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,3584,0.00979733301533593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,8192,2560,0.009902222288979424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,8192,3072,0.018564444449212816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,3072,0.008973333570692275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,3072,0.14433333608839247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,8192,8192,0.02368800010946062
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,8192,2048,0.008857777549160851
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,2560,0.008618666893906063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,2560,0.14132444063822427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,8192,1536,0.007271111011505127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,10240,0.14012977812025282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,8192,2048,0.015332443846596612
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,2048,0.1443342235353258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,2048,0.007997333175606197
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,8192,1024,0.005230222311284807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,8192,1536,0.015046222342385186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,1536,0.007255111303594377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,8192,768,0.004156444635656145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,8192,1024,0.01239911135700014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,1024,0.006958222223652734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,1024,0.14485332700941297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,8192,512,0.0036115555299652945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,768,0.006486222147941589
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,8192,768,0.012717333104875354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,768,0.14095732900831434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,8192,256,0.003177777760558658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,8192,512,0.012070222033394707
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,512,0.1413608921898736
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,512,0.006243555496136348
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,8192,128,0.002891555635465516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,8192,256,0.011385777758227454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,256,0.005917333480384614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,256,0.1448977788289388
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,8192,64,0.0028560000161329904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,8192,128,0.011679111255539788
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,8192,32,0.0028577778074476454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,8192,128,0.005903111149867375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,128,0.13420977857377794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,8192,64,0.011090666883521609
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,8192,32,0.010024000373151567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,7168,65536,0.14751822418636745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,7168,16384,0.043024000194337636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,7168,65536,0.09264800283643936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,65536,0.07407377825842963
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,16384,0.022468444373872545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,7168,12288,0.032608889871173434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,16384,0.14123733838399252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,8192,1536,0.1386302179760403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,7168,12288,0.03458399905098809
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,7168,10240,0.029127111037572224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,12288,0.14407911565568712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,12288,0.018552000323931377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,65536,0.14944533507029215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,7168,8192,0.02514755560292138
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,7168,10240,0.029905776182810467
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,10240,0.016522667474216886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,8192,4096,0.019329778022236295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,10240,0.1448062260945638
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,7168,7168,0.018988443745507132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,7168,8192,0.02693866689999898
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,8192,0.014198222094111972
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,8192,0.1422791083653768
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,7168,6144,0.016662221815851
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,7168,7168,0.024384000235133704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,7168,0.013440888788965015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,7168,0.14642311467064753
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,7168,5120,0.0144177774588267
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,7168,6144,0.023132445083724126
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,6144,0.012308444413873883
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,8192,2560,0.01718311177359687
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,7168,5120,0.02070844504568312
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,5120,0.14185955789354113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,7168,16384,0.043635557095209755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,7168,4096,0.019241778386963736
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,7168,3584,0.011043555206722684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,4096,0.010002666877375709
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,4096,0.14581333266364202
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,7168,3072,0.010038221875826517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,7168,3584,0.01868888901339637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,3584,0.009297777381208207
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,3584,0.14522400167253283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,7168,2560,0.009331555830107795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,7168,3072,0.017678222722477384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,3072,0.008275555239783393
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,3072,0.13975021574232313
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,7168,2048,0.007621333003044128
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,7168,2560,0.01685688893000285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,2560,0.007927999728255803
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,2560,0.14653421772850886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,7168,1536,0.006640000061856375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,7168,2048,0.015264888604482016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,2048,0.007292444507280986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,7168,1536,0.014115555418862237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,7168,1024,0.005568888866239124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,1536,0.14008622699313694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,1536,0.006931555353932911
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,6144,0.13908621999952528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,7168,1024,0.012131555212868584
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,1024,0.006506666541099548
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,7168,768,0.003980444537268745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,7168,4096,0.01236355553070704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,1024,0.1396604379018148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,7168,512,0.003551111039188173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,5120,0.011374221907721626
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,768,0.006196444647179709
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,7168,768,0.012087999946541257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,7168,512,0.011697777443461947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,512,0.1440017753177219
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,512,0.005906666732496685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,7168,128,0.0028151110228565005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,7168,256,0.010417777631017896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,256,0.1403635607825385
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,256,0.005909333212508096
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,7168,64,0.002543111021320025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,7168,128,0.011696889168686338
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,128,0.1375395589404636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,7168,128,0.0058755555914507965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,7168,64,0.011081777513027191
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,7168,32,0.009686222506894005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,6144,65536,0.12415911091698541
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,6144,65536,0.0848275555504693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,6144,16384,0.03838222225507101
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,65536,0.07200533151626587
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,6144,16384,0.03804088963402642
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,6144,12288,0.030167112747828167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,16384,0.02230666743384467
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,16384,0.13932977782355413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,7168,256,0.0031431110368834603
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,65536,0.14910577403174505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,6144,10240,0.02544266647762722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,12288,0.01794755624400245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,12288,0.14247822761535645
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,6144,8192,0.02186133298608992
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,6144,10240,0.028379556205537584
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,10240,0.1390684445699056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,7168,32,0.0028124445428450904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,6144,8192,0.025674665967623394
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,6144,7168,0.019706666469573975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,8192,0.01418133411142561
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,8192,0.14643822775946722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,6144,6144,0.017704889178276062
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,6144,7168,0.023584889041052923
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,7168,0.1450737714767456
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,2048,0.14477154943678114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,6144,6144,0.022638221581776936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,6144,5120,0.015971556305885315
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,6144,0.14103733168707952
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,6144,5120,0.020213334096802604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,5120,0.14567911624908447
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,6144,4096,0.011448889142937131
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,5120,0.010743111371994019
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,6144,4096,0.01891377733813392
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,7168,768,0.1417093276977539
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,6144,3584,0.010429333481523724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,4096,0.14676000012291804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,4096,0.009586667021115622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,6144,3584,0.018458666072951425
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,10240,0.016018667154841952
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,6144,3072,0.009503111243247986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,3584,0.1452382273144192
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,3584,0.00889511075284746
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,6144,2560,0.008377778033415476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,6144,3072,0.01688888834582435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,3072,0.00831288927131229
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,3072,0.14347022109561497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,6144,2560,0.01638488968213399
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,6144,2048,0.007592889169851939
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,2560,0.007576888634098901
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,2560,0.14115911059909397
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,6144,1536,0.006271111054552927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,6144,2048,0.014449778530332776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,2048,0.14059466785854763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,2048,0.0069360000391801195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,6144,0.012025777664449481
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,6144,1024,0.003927999900446998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,6144,1536,0.013769778112570444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,1536,0.006955555743641323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,1536,0.1446266704135471
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,6144,768,0.0038399998512532976
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,6144,1024,0.012378666963842181
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,1024,0.006251555350091722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,1024,0.14439822567833796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,6144,512,0.0033235556135574975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,6144,768,0.012086222569147745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,768,0.006256000035338932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,6144,256,0.003160888950030009
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,6144,12288,0.03596355517705282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,6144,512,0.010759111079904767
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,512,0.14045333862304688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,6144,128,0.0028248888750871024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,6144,256,0.011024889018800525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,256,0.005636444522274866
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,256,0.1412408881717258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,6144,64,0.002472888885272874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,6144,128,0.011702222128709158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,6144,32,0.0028293333533737394
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,128,0.13572888904147679
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,6144,64,0.011176889141400656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,6144,32,0.010371555884679159
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,5120,65536,0.10746133327484131
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,7168,0.013050666285885705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,5120,65536,0.07624799675411649
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,5120,16384,0.03346488873163859
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,65536,0.06930844651328193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,5120,16384,0.034874667723973594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,5120,12288,0.02731200059254964
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,16384,0.1387831105126275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,16384,0.021024000313546922
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,65536,0.14931466844346789
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,6144,768,0.14097511768341064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,5120,12288,0.03283910950024923
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,12288,0.017278222574128043
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,5120,10240,0.023003555006451074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,12288,0.14055821630689833
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,5120,8192,0.020640888147883944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,5120,10240,0.027898665931489732
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,10240,0.015490666031837463
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,10240,0.14172177844577366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,5120,8192,0.024777778320842322
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,5120,7168,0.01791111131509145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,8192,0.14211377832624647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,8192,0.013731555806265937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,128,0.005598222215970357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,5120,7168,0.02295644415749444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,7168,0.13906666967603895
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,7168,0.012607999973826937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,5120,5120,0.013988445202509562
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,6144,0.011454222102959951
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,6144,0.14789599842495388
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,5120,4096,0.010370666782061258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,5120,5120,0.01969688965214623
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,5120,0.01016444464524587
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,5120,4096,0.018202667435010273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,5120,3584,0.009639110830095079
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,4096,0.00849599970711602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,5120,3584,0.017228444417317707
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,3584,0.14084445105658636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,5120,3072,0.008435555630260045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,3584,0.008040889269775814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,6144,512,0.005977777971161737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,5120,3072,0.01715022159947289
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,5120,2560,0.007924444145626491
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,3072,0.1464782158533732
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,3072,0.008017777568764156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,5120,2048,0.0069511110583941145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,5120,2560,0.015394666128688388
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,5120,6144,0.016153777639071148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,2560,0.1460177765952216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,2560,0.007655111451943715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,5120,1536,0.006255110932721033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,5120,2048,0.014094221923086377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,5120,6144,0.021930666433440313
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,2048,0.14085066318511963
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,2048,0.006940444310506185
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,5120,1024,0.004515555583768421
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,5120,1536,0.013760888742076026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,1536,0.006919999917348226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,5120,0.13961422443389893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,5120,768,0.0035546666218174826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,5120,1024,0.012076444096035428
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,1024,0.14215732945336237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,5120,512,0.003488000068399641
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,4096,0.14789511097802058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,768,0.1415004465315077
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,5120,512,0.011683555940786997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,5120,256,0.0031439999325407874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,512,0.14390577210320368
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,5120,256,0.011335999601417117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,256,0.1433431175020006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,5120,128,0.002893333426780171
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,256,0.005589333259397083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,5120,128,0.011624000138706632
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,128,0.1340142223570082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,128,0.005624889085690181
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,5120,64,0.0024871110088295406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,5120,32,0.002514666567246119
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,5120,64,0.011200000014570026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,5120,32,0.010363555616802638
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,5120,1536,0.13908178276485866
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,4096,65536,0.08953866693708633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,1024,0.006424888968467712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,4096,16384,0.027868443065219458
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,5120,768,0.012390221986505719
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,4096,65536,0.06900799936718412
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,65536,0.06726400057474773
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,16384,0.02054933375782437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,4096,12288,0.021743999587164983
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,65536,0.1500284406873915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,4096,12288,0.027779555983013574
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,4096,10240,0.018558222386572096
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,12288,0.14013955328199598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,12288,0.01680088871055179
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,4096,10240,0.02609066665172577
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,10240,0.13950222068362766
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,4096,8192,0.01605599953068627
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,10240,0.015099555253982544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,4096,8192,0.023581332630581323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,8192,0.013048888908492194
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,8192,0.14632444911532932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,4096,7168,0.015081778168678284
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,7168,0.012083555261294046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,7168,0.14413689242468938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,4096,6144,0.013784888717863293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,768,0.005919999960396025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,4096,6144,0.020450666546821594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,6144,0.1406008932325575
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,4096,16384,0.03513955407672458
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,16384,0.13985955715179443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,4096,5120,0.019031110737058852
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,4096,4096,0.010994666980372535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,5120,0.009320889082219865
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,4096,4096,0.017489777670966256
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,4096,0.13743733035193548
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,4096,3584,0.010995555255148145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,4096,0.008953777452309927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,4096,3584,0.01646311084429423
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,3584,0.13987199465433756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,4096,3072,0.010782221953074137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,3584,0.007986666427718269
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,4096,7168,0.021571555071406897
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,4096,3072,0.016021332807011075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,3072,0.13994667265150282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,4096,2560,0.009122666385438707
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,3072,0.007653333246707916
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,6144,0.011325333681371478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,4096,2560,0.014074666632546319
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,4096,5120,0.01220533334546619
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,4096,2048,0.00812888890504837
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,2560,0.007306666837798224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,4096,1536,0.005917333480384614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,4096,2048,0.01368533323208491
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,2048,0.006956444846259222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,5120,0.14505422115325928
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,4096,1024,0.005173333403136996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,1536,0.006615110983451207
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,1536,0.13901066780090332
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,4096,1024,0.0124195557501581
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,4096,768,0.0042462220622433555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,1024,0.006291555447710886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,1024,0.14531821674770778
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,4096,512,0.003866666720973121
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,4096,768,0.011767110890812345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,768,0.14021955596076116
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,768,0.006232888748248418
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,4096,256,0.0028684443483750024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,4096,512,0.01070577816830741
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,512,0.14470844798617893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,4096,256,0.011389333340856763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,256,0.14499555693732366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,4096,128,0.0028524444335036804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,256,0.005902222047249476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,5120,512,0.005976000179847081
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,2560,0.13962133725484213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,128,0.13163467248280844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,4096,64,0.0025902222841978073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,4096,32,0.0028168888141711554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,128,0.005610666755172942
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,4096,2048,0.14284800158606634
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,4096,1536,0.013400000002649097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,4096,32,0.010327999790509542
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3584,65536,0.06613955232832167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3584,16384,0.023795555035273235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,65536,0.06603911187913683
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3584,16384,0.03018933203485277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,65536,0.14827822314368352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,16384,0.1411457724041409
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,4096,512,0.005903111149867375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3584,12288,0.019634667370054457
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,16384,0.020233778489960563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3584,10240,0.017313778400421143
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3584,12288,0.029831111431121826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,12288,0.016626665989557903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,12288,0.1425066656536526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3584,8192,0.015103111664454142
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3584,10240,0.023988443944189284
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,10240,0.014892444014549255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,4096,128,0.010805333654085795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3584,7168,0.014148443937301636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,8192,0.012729778057999082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,8192,0.14351022243499756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,4096,64,0.011022222538789114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3584,7168,0.021213332811991375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3584,6144,0.012404444317022959
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3584,65536,0.08119022183948092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,7168,0.011704888608720569
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,7168,0.13994133472442627
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3584,5120,0.011704888608720569
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3584,6144,0.01994044416480594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,6144,0.010668444136778513
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3584,4096,0.010326222413116032
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3584,5120,0.018771555688646104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,5120,0.14504888322618273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,5120,0.009371555513805812
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3584,3584,0.010230222509966956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3584,4096,0.01718933383623759
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,4096,0.14024355676439074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,4096,0.008338666624493068
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3584,3072,0.009646221995353699
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3584,3584,0.014565333724021912
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,3584,0.007961778177155389
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,3584,0.14180266857147217
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3584,2560,0.008897778060701158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3584,3072,0.014831999937693277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,3072,0.007640000018808577
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3584,2560,0.014858666393491956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,2560,0.14339289400312635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,2560,0.007298666569921706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,10240,0.14013955328199598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3584,8192,0.021472889516088698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3584,1536,0.005964444329341252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3584,2048,0.01312800000111262
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,2048,0.13951022095150417
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3584,1536,0.012727999852763282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,1536,0.0069520001610120135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3584,1024,0.004858666823969947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,6144,0.13897600438859728
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,1024,0.006267555471923616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3584,1024,0.012410666379663678
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,1024,0.14547733465830484
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3584,768,0.012084444363911947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,768,0.006278222219811545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,768,0.14011732737223306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3584,512,0.011733333269755045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,512,0.13948444525400797
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3584,256,0.0028693332440323303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,3072,0.1460968918270535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,512,0.00591466658645206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3584,256,0.011056000159846412
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3584,2048,0.007700444095664554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3584,128,0.0028106667515304354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,256,0.00574933323595259
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,256,0.14652889304690891
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3584,64,0.0028533333291610083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3584,128,0.01203644441233741
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,128,0.005748444547255834
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,3584,2048,0.0069217777086628815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,128,0.1328835619820489
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3584,64,0.010078222387366826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3072,65536,0.069504890176985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,3584,1536,0.1455333365334405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3584,32,0.009352000223265754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3072,16384,0.020942222740915086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3072,65536,0.06361866659588285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3584,768,0.004211555752489301
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3072,16384,0.030493332280053034
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,65536,0.06530755758285522
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,16384,0.020210666788948905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,16384,0.14540178245968288
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,65536,0.14972621864742702
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3584,512,0.0038648889296584656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3072,12288,0.025192888246642217
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3072,10240,0.015578667322794596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,12288,0.01632355484697554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3072,8192,0.01367111090156767
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3072,10240,0.023541332946883306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,10240,0.014462222655614218
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3072,8192,0.021360889077186584
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3072,7168,0.01275644451379776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,8192,0.012274666792816587
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3072,7168,0.019927110936906602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,7168,0.14776621924506292
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3072,6144,0.012058667010731168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,7168,0.010816000401973724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3072,6144,0.018889778190188937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,6144,0.14366666475931802
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3072,5120,0.010688888529936472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,6144,0.010028444230556488
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3072,12288,0.017998221847746108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3072,5120,0.017842666970358956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,5120,0.14735466904110378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,5120,0.009296000003814697
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3072,4096,0.010026666853162978
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,12288,0.14131200313568115
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3072,3584,0.009358222285906473
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3072,4096,0.015488000379668342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,4096,0.00833422193924586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,4096,0.1380231115553114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,10240,0.14617333147260878
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3072,3072,0.008811555802822113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3072,3584,0.014476444986131458
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,3584,0.00796088907453749
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3072,2560,0.007874666816658443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,3584,0.1443271107143826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3072,3072,0.014105778601434497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,3072,0.007649777664078607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,3072,0.13924445046318903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3072,2048,0.006554666492674086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3072,2560,0.013470222552617392
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,2560,0.007612444460391998
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,2560,0.1452817784415351
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3072,1536,0.005904888941182031
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3072,2048,0.013201778133710226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3584,32,0.0028115556471877625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,2048,0.13899200492435032
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,2048,0.007265778051482306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3072,1024,0.004884444591071871
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3072,1536,0.012406222522258759
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,1536,0.006907555378145642
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,1536,0.1449857817755805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,1024,0.006251555350091722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,1024,0.143504884507921
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3072,512,0.003900444342030419
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3072,768,0.012063999970753988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,768,0.006133333262470033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,768,0.1449671056535509
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3072,256,0.003522666792074839
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3072,512,0.011686222420798408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,512,0.14010932710435656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3072,256,0.01143200033240848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3072,128,0.0031715554909573663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,256,0.005588444570700328
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,256,0.1446897718641493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3072,128,0.011723555624485016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,128,0.005573333137565189
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3072,64,0.0028444443725877335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3072,32,0.0024951110697454875
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,8192,0.14170311556922063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3072,64,0.011030221978823343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2560,65536,0.05907377931806776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3072,32,0.009696889254781935
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2560,65536,0.061378664440578885
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,65536,0.06438577837414212
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,3072,768,0.004552888787455029
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,3072,1024,0.012077333198653327
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,65536,0.1572879950205485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2560,16384,0.02959111001756456
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2560,12288,0.016063110695944894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,16384,0.14686311615837946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,16384,0.01984177695380317
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2560,10240,0.013743110828929476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2560,12288,0.024069334069887798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,3072,512,0.005961777849329843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,12288,0.015607999430762397
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,12288,0.13944355646769205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2560,8192,0.01256533298227522
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2560,10240,0.02256711158487532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,10240,0.014158222410413953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2560,7168,0.011548444628715515
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2560,8192,0.019549333386951022
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,8192,0.012093333734406365
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,8192,0.14747022257910833
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,3072,128,0.13223111629486084
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2560,6144,0.01074044406414032
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2560,7168,0.01958755486541324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,7168,0.010675555302037133
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2560,5120,0.009908444351620143
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2560,6144,0.017839110559887357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,6144,0.00999466660949919
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,6144,0.14662844604916042
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2560,4096,0.009386666946940953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2560,5120,0.01590222285853492
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,5120,0.14092089070214167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,5120,0.009358222285906473
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2560,3584,0.008818666968080733
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2560,16384,0.01958488921324412
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2560,4096,0.01482488877243466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,4096,0.00831733312871721
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,4096,0.14106755786471897
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2560,3072,0.008650666309727563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2560,3584,0.014456889695591398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,3584,0.00796977761718962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2560,2560,0.006894222150246303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2560,3072,0.014127999544143677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,3072,0.007651555869314406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,3072,0.1400551133685642
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2560,2048,0.005902222047249476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2560,2560,0.01310400002532535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,2560,0.14103378189934626
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,2560,0.007300444775157505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2560,1536,0.00657155571712388
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2560,2048,0.013058666553762225
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,10240,0.14526844024658203
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,2048,0.0069022224181228215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,2048,0.1471573379304674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2560,1024,0.004868444469239977
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2560,1536,0.012721777790122561
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,1536,0.0069431112044387394
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,1536,0.13894755310482448
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2560,768,0.0042151109211974675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2560,1024,0.011710222396585675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,1024,0.006303999986913469
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,7168,0.14044445090823704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2560,512,0.003663111064169142
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2560,768,0.012006222373909421
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,768,0.005940444353553984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,768,0.14626488420698378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2560,256,0.0032213332338465583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2560,512,0.011428444749779172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,512,0.14105244477589926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,512,0.005932444499598608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2560,128,0.0028444443725877335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2560,256,0.011476444701353708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,256,0.005599110904667113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,256,0.14594488673739964
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2560,64,0.002540444541308615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2560,128,0.011744000017642975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2560,32,0.002838222309947014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,2560,128,0.0058240000572469496
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,128,0.13437867164611816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,3584,0.1460746659172906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2560,64,0.01016622202263938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2560,32,0.01032266683048672
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2048,16384,0.016574222180578444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2048,65536,0.059142225318484835
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,65536,0.061872886286841504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2048,16384,0.025264889001846313
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,16384,0.14704710907406277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,65536,0.14968621730804443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,16384,0.01920711166328854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2048,10240,0.011730666789743634
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2048,12288,0.022895110978020564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,12288,0.015487111277050443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,12288,0.1467591126759847
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2048,10240,0.020474667350451153
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,10240,0.012920888761679331
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,10240,0.14003554979960123
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,2560,1024,0.13970755206214056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2048,7168,0.010971555279360877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2048,8192,0.01890222231547038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,8192,0.011353777514563667
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,8192,0.1408133374320136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2048,6144,0.009737778041097853
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2048,7168,0.017487110363112558
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,7168,0.010475555227862464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2048,6144,0.016844444804721408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2048,5120,0.009000000026490953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,6144,0.14183378219604492
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,6144,0.009931555224789513
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2048,65536,0.05208622084723579
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2048,5120,0.015096889601813423
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,5120,0.00923288861910502
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,5120,0.13925688796573216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2048,4096,0.014476444986131458
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,4096,0.008323555191357931
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,4096,0.14169243971506754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2048,3584,0.009337777892748514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,3584,0.007991111112965478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2048,3072,0.0069066666894488865
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,3584,0.13873688379923502
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2048,8192,0.011385777758227454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2048,3072,0.013770666387346057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,3072,0.13763288656870523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,3072,0.007628444168302748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2048,2560,0.013397333522637686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,2560,0.14606132772233751
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2048,2048,0.006958222223652734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,7168,0.14567200342814127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,2560,0.007249777515729268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2048,2048,0.012415111064910889
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,2048,0.13622132937113443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2048,4096,0.009694221946928237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2048,1536,0.006028444402747684
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,2048,0.0069573331210348345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,1536,0.00666222224632899
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2048,1536,0.012423111332787408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,1536,0.14011555247836643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2048,12288,0.013074666261672974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2048,1024,0.011781333221329583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2048,3584,0.013910222384664746
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,1024,0.006252444452709622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,1024,0.13929333951738146
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2048,768,0.012030222349696688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,768,0.005958222266700532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,768,0.14349155955844456
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2048,2560,0.006055111272467508
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2048,256,0.0031964443624019623
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2048,512,0.011119999819331698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,512,0.00589155571328269
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2048,256,0.011408000356621213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2048,128,0.0028479999552170434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,256,0.005574222240183089
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,256,0.14044445090823704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2048,128,0.012085333466529846
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,2048,128,0.005601777798599667
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,128,0.1370240052541097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2048,32,0.0028071111689011254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2048,64,0.010880000061459012
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,2048,32,0.010048000348938836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1536,65536,0.0373582210805681
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2048,1024,0.004872888740566042
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1536,65536,0.057335999276902944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1536,16384,0.013795555465751223
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2048,768,0.003900444342030419
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,65536,0.060079110993279346
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,65536,0.14792533715566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1536,12288,0.011359111302428775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2048,512,0.0038382220599386427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1536,16384,0.023578666978412207
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,16384,0.018269333574506972
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,16384,0.14186399512820774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1536,10240,0.01036088913679123
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1536,12288,0.021670222282409668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,12288,0.013982221484184265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,12288,0.14027644528283015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,2048,512,0.138208892610338
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1536,10240,0.01923644377125634
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1536,8192,0.01019200020366245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1536,7168,0.009649777577983009
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,10240,0.012757333616415659
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,10240,0.14526400301191542
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1536,8192,0.016566221912701923
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,8192,0.011183111204041375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,8192,0.13841688632965088
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1536,6144,0.008935111264387766
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1536,7168,0.015799999237060547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,7168,0.14020533031887478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,2048,64,0.002846222163902389
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,6144,0.010010666317409938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,6144,0.1379635598924425
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1536,4096,0.00778311077091429
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1536,5120,0.014867555763986377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,5120,0.009266667068004608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,5120,0.1412551138136122
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1536,3584,0.007306666837798224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1536,4096,0.014143110977278816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,4096,0.008265777594513362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,4096,0.13954044712914362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1536,3072,0.006263111200597551
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1536,3584,0.013789333403110504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,3584,0.007976888782448238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,3584,0.1363662216398451
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1536,2560,0.005532444351249271
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,3072,0.007629333270920648
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,3072,0.1455004480150011
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1536,2560,0.013221333424250284
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,2560,0.007258666886223688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,2560,0.13871199554867214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1536,2048,0.012741333080662621
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,2048,0.006970666348934174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,2048,0.14140000608232287
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1536,1536,0.01238044434123569
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,7168,0.010354667074150508
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1536,1024,0.004327110946178436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,1536,0.006614222294754452
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1536,6144,0.015351111690203348
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1536,1024,0.011709333293967776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1536,768,0.003879111260175705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,1024,0.1404328876071506
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1536,768,0.011385777758227454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,768,0.13802399900224474
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1536,512,0.0034906665484110513
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,768,0.005935110979610019
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1536,3072,0.013741333451535968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1536,512,0.011384888655609555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1536,2048,0.006272000157170826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,512,0.14630756113264295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,512,0.005884444548024072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1536,256,0.011394666300879585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1536,1536,0.005180444568395615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,256,0.14439733823140463
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1536,128,0.002850666642189026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,256,0.005618666609128316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1536,128,0.01074488874938753
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,128,0.13346044222513834
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,128,0.005601777798599667
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1536,32,0.0028417776856157514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1536,64,0.01105955574247572
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1536,32,0.009657777845859528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,1536,1024,0.0062977779242727495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1024,65536,0.027248000105222065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1024,65536,0.05530044436454773
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,65536,0.05842577748828464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1024,16384,0.01462222304609087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1024,16384,0.022269333402315777
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,16384,0.1455155611038208
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1536,256,0.00295733329322603
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1024,12288,0.0120666664507654
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,16384,0.01681777834892273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1024,12288,0.01936444474591149
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,12288,0.1464737786187066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,12288,0.013958222336239286
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1024,10240,0.017702221870422363
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,1536,1536,0.14159822463989258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1536,64,0.0026088888860411118
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1024,8192,0.010989333192507425
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,10240,0.01277511070171992
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1024,8192,0.016176000237464905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,8192,0.14141066869099936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1024,7168,0.009394666386975182
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1024,7168,0.016171556380059984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,7168,0.14673244953155518
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,65536,0.1476622157626682
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1024,6144,0.008721777962313758
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,7168,0.010446222292052375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1024,6144,0.015140444040298462
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,6144,0.009702222214804756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1024,5120,0.007987555530336168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1024,5120,0.014820444915029736
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,5120,0.1407911115222507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1024,10240,0.010747555229398938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,5120,0.009309333231714036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1024,4096,0.013745778136783175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,10240,0.13932889037662083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1536,5120,0.007400888535711501
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,4096,0.14176621701982287
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,4096,0.008329778081840938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1024,3584,0.013493333425786761
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,3584,0.14252532853020564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,8192,0.011385777758227454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,3584,0.00796977761718962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1024,3072,0.013750221994188098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1024,2560,0.0058871110280354815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1024,2560,0.013140444126394061
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,6144,0.14600800143347845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,2560,0.14124266306559244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,2560,0.00720266666677263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1024,4096,0.007325333025720384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1024,2048,0.012083555261294046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,2048,0.14057332939571804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,2048,0.006618666566080517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1024,1536,0.005260444349712796
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1024,1536,0.011872889267073737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1024,1024,0.0046453335218959385
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1024,3072,0.006238222122192383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,1536,0.00674311113026407
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1024,1024,0.011706666813956367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,1024,0.006265777680608962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1024,768,0.004231111043029361
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,1024,0.14688889185587564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1024,768,0.011640888949235281
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,768,0.14654578102959528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1024,512,0.0038702223036024305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1024,512,0.010406222608354356
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1024,2048,0.006567111031876669
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1024,256,0.003145777723855443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,512,0.005912000106440649
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1024,256,0.01130844404300054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,256,0.14623110824161106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1024,3584,0.006903999795516332
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1024,128,0.011776000261306763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,1536,0.14018221696217856
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,128,0.005602666487296422
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,128,0.13153421878814697
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1024,64,0.002536000063021978
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,3072,0.145778669251336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1024,32,0.002817777709828483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,768,65536,0.021684444612926904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1024,64,0.011035555766688453
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,768,0.005941333456171884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,768,65536,0.05305599835183886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,768,16384,0.01033688916100396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,65536,0.05818044477038913
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,65536,0.14728800455729166
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,1024,512,0.14165600140889487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,768,12288,0.011172444456153445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,768,16384,0.02097333305411869
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,16384,0.016768889294730294
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,16384,0.14494400554233128
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,768,12288,0.018901333212852478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,768,10240,0.009672889278994666
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,12288,0.1417751047346327
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,12288,0.013788444300492605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,256,0.005596444424655702
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,768,8192,0.009648000200589498
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,10240,0.012749333348539142
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,768,10240,0.017816889617178176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,10240,0.14167643917931452
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,768,7168,0.008960000342792934
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,768,8192,0.01648000048266517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,8192,0.011338666909270816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,8192,0.14517866240607366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,768,6144,0.007988444632954067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,768,7168,0.015806222955385845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,7168,0.14245600170559353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,7168,0.01037688884470198
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,768,5120,0.006894222150246303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,768,6144,0.015119999647140503
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,6144,0.14169689019521078
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,1024,32,0.009341333475377824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,768,4096,0.007280000381999546
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,5120,0.00902844468752543
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,5120,0.14449332820044622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,768,4096,0.014121777481502958
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,768,3584,0.007292444507280986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,4096,0.14090666506025526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,768,3584,0.01384622189733717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,3584,0.139284438557095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,768,3072,0.006628444625271692
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,1024,128,0.0028435554769304064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,768,3072,0.013471110827393003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,3072,0.14743822150760227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,768,2560,0.006594666590293248
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,3072,0.007647111184067196
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,1024,3072,0.007600888609886169
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,768,2560,0.012412444584899478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,2560,0.14020000563727483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,768,2048,0.005224888937340842
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,2560,0.007354666789372762
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,6144,0.009932444327407414
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,768,1536,0.004914666629499859
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,768,5120,0.014828445182906257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,768,2048,0.012442666623327466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,2048,0.14553244908650717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,768,1024,0.0041999999019834726
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,768,1536,0.012387555506494312
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,1536,0.006591111007663939
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,1536,0.14033333460489908
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,4096,0.008279999925030602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,768,768,0.003826666623353958
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,768,1024,0.011732444167137146
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,1024,0.006280888699822956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,1024,0.14082133769989014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,768,512,0.0031999999450312722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,768,768,0.011755555868148804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,768,0.13811288939581975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,768,256,0.0028533333291610083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,768,512,0.011360000405046674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,3584,0.008024888734022776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,512,0.14034667279985216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,512,0.005630222045713001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,768,128,0.0028151110228565005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,768,256,0.011050666371981302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,256,0.005616888817813661
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,256,0.1416648891237047
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,768,128,0.011708444191349877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,768,64,0.0028373334142896864
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,768,32,0.0026488889836602737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,128,0.005589333259397083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,512,65536,0.017535999417304993
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,768,64,0.011063999599880643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,768,32,0.009709333380063375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,512,16384,0.010770666930410596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,512,65536,0.05262666609552172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,65536,0.14602577686309814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,65536,0.05662844578425089
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,512,12288,0.008743111458089616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,512,16384,0.02164177762137519
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,16384,0.14076622327168783
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,2048,0.00692622239391009
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,512,10240,0.008301333420806462
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,512,12288,0.018912000788582694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,12288,0.14670844872792563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,512,10240,0.01755199995305803
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,10240,0.1400737762451172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,10240,0.012763555679056378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,768,768,0.0059279998143513995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,512,8192,0.016536000702116225
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,8192,0.14609866672092015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,8192,0.011039999624093374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,512,7168,0.01571377780702379
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,7168,0.146598219871521
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,768,128,0.13262932830386692
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,512,6144,0.007075555622577667
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,7168,0.010480889015727572
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,512,6144,0.015133332875039844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,6144,0.14568977885776097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,512,5120,0.006962666908899943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,6144,0.010021333065297868
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,512,5120,0.014116444521480136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,16384,0.01682222220632765
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,5120,0.14527111583285862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,512,4096,0.006962666908899943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,5120,0.00904977818330129
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,12288,0.014074666632546319
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,512,3584,0.006924444602595435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,512,4096,0.014138667119873894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,4096,0.008351999852392409
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,4096,0.13957156075371638
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,512,3072,0.006588444527652528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,512,8192,0.007999999655617608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,512,3584,0.013798221945762634
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,3584,0.007985778152942657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,3584,0.14675111240810818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,512,3072,0.013447999954223633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,3072,0.007651555869314406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,3072,0.14436622460683188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,512,2048,0.005927111125654644
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,512,7168,0.007640000018808577
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,512,2560,0.012813333008024426
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,2560,0.14569066630469427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,2560,0.007261333366235097
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,512,2048,0.012433778080675336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,2048,0.006912888752089606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,2048,0.13928622669643825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,512,1024,0.0038382220599386427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,512,1536,0.011447110937701331
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,1536,0.13834577136569554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,512,768,0.003536888708670934
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,512,1024,0.011742221812407175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,1024,0.006244444598754247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,512,768,0.011705777711338468
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,768,0.005941333456171884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,512,512,0.0034888887570963967
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,512,512,0.01035911093155543
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,512,0.13943644364674887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,512,256,0.003145777723855443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,512,0.00564177789621883
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,512,256,0.01107022249036365
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,512,2560,0.006271111054552927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,256,0.1471760008070204
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,512,128,0.0028231110837724474
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,256,0.005552000055710475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,512,128,0.010596444209416708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,512,1536,0.004378666480382283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,128,0.130876448419359
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,512,64,0.0028275555620590844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,512,32,0.0027831111931138565
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,256,65536,0.01605866683853997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,512,64,0.010370666782061258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,512,32,0.009036444127559662
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,1536,0.006629333313968446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,256,16384,0.009014222357008193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,1024,0.14060533046722412
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,256,65536,0.049858666128582425
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,65536,0.05064622229999966
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,256,12288,0.008314666648705801
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,256,16384,0.02085599965519375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,16384,0.01683911184469859
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,16384,0.14274311065673828
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,256,10240,0.007592889169851939
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,256,12288,0.018234666850831773
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,12288,0.014064888159434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,256,10240,0.017887999614079792
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,256,8192,0.006968888971540663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,10240,0.14513688617282444
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,256,8192,0.01650311052799225
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,8192,0.14611111746893987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,256,7168,0.006602666858169768
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,8192,0.011063111325105032
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,512,128,0.005605333381228977
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,256,7168,0.015388444066047668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,256,6144,0.006603555546866522
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,7168,0.14311911000145808
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,7168,0.010381333529949188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,256,6144,0.014787554740905762
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,256,5120,0.005984888722499211
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,6144,0.009680888719028896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,65536,0.14790666103363037
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,512,768,0.13967466354370117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,256,5120,0.013767999907334646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,5120,0.14039733674791124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,256,4096,0.0058746664888328975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,12288,0.14611021677652994
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,256,4096,0.013101333545313941
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,256,3584,0.006021333237489064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,4096,0.00831288927131229
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,4096,0.13852532704671225
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,256,3072,0.006228444476922353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,256,3584,0.012753778033786349
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,3584,0.14390399720933703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,256,3072,0.013752000199423896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,256,2560,0.006248000181383557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,3072,0.007669332954618666
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,3072,0.14619733227623835
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,256,2560,0.013071111506885953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,2560,0.007167999943097432
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,256,2048,0.006060444646411472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,2560,0.14424622058868408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,256,2048,0.011744000017642975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,2048,0.14561333921220568
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,256,1536,0.004512000001139111
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,2048,0.006602666858169768
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,256,1536,0.012381333443853589
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,1536,0.006617777877383762
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,256,1024,0.0038542221817705366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,1536,0.14381778240203857
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,1024,0.0063075555695427795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,256,1024,0.01110400011142095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,1024,0.14111910925971136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,256,768,0.003504000190231535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,10240,0.012711111042234631
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,256,512,0.0031475555151700974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,3584,0.007939555578761632
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,256,768,0.011071999867757162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,768,0.006181333214044571
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,768,0.1405537790722317
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,256,256,0.0028231110837724474
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,512,0.005757333503829108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,256,512,0.011304000185595619
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,256,128,0.0027955555253558685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,256,256,0.011381333072980246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,256,0.1426266696718004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,256,0.005576000031497743
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,256,64,0.002518222149875429
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,256,128,0.011590222517649332
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,128,0.005609777652555042
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,128,0.13228355513678655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,256,32,0.002471999989615546
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,128,65536,0.01443377799457974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,256,64,0.010049777726332346
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,256,32,0.010033778018421596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,128,16384,0.007644444704055786
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,128,65536,0.04894222153557671
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,6144,0.14135111702813044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,65536,0.14632977379692927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,65536,0.04961688982115852
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,128,12288,0.007270221908887227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,128,16384,0.020652444826232064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,16384,0.01683466633160909
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,256,5120,0.009312888814343346
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,128,12288,0.01811644434928894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,128,10240,0.006948444578382704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,12288,0.1415919992658827
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,12288,0.013815999858909182
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,128,8192,0.006752000086837345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,10240,0.012701333396964602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,128,10240,0.017533333765135873
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,10240,0.1473146677017212
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,128,7168,0.006611555400821898
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,128,8192,0.01587199999226464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,8192,0.14004266262054443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,8192,0.011053333679835001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,128,6144,0.005940444353553984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,128,7168,0.015300444430775113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,7168,0.010392888552612728
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,7168,0.1471502251095242
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,128,5120,0.005886222339338726
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,128,6144,0.015485333071814643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,256,512,0.13996622297498915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,6144,0.009711110757456886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,6144,0.14105689525604248
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,128,4096,0.00591555568906996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,128,5120,0.01479288935661316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,5120,0.009003555609120263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,128,3584,0.005924444645643234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,5120,0.14558755026923284
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,128,4096,0.01313866674900055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,4096,0.008289777570300633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,4096,0.14090932740105525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,128,3072,0.005589333259397083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,128,3584,0.013760888742076026
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,3584,0.00794400026400884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,128,2560,0.006278222219811545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,128,3072,0.013451555536852943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,3072,0.007343110938866933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,3072,0.1398808823691474
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,128,2048,0.005599110904667113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,128,2560,0.012726222475369772
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,2560,0.007291555404663086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,2560,0.14153422249688044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,128,1536,0.0042160000238153665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,128,2048,0.012545777691735161
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,16384,0.1395662228266398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,2048,0.0069777775141927935
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,2048,0.14723732736375597
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,128,1024,0.0038382220599386427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,128,1536,0.012438221938080259
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,1536,0.006623999940024481
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,1536,0.14051022794511583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,128,768,0.003543111185232798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,128,1024,0.011759111450778114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,1024,0.0062773335311147906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,1024,0.14502488242255315
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,128,512,0.0031439999325407874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,768,0.005950222412745158
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,768,0.13935466607411703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,128,512,0.011046222514576383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,512,0.0058959999846087555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,512,0.14590844843122694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,128,256,0.011008000208271874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,128,128,0.0028168888141711554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,256,0.1428826649983724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,128,128,0.011696000066068439
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,128,0.13901510503556994
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,128,64,0.002509333400262727
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,128,0.005599110904667113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8_block,4,128,3584,0.144813338915507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,128,64,0.010716444088353051
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,64,65536,0.010715555813577441
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,128,32,0.009688888986905416
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,64,16384,0.007280888656775157
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,64,65536,0.04823466804292467
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,64,12288,0.006299555715587404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,64,16384,0.021013332737816706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,64,10240,0.006256000035338932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,128,768,0.012039110892348819
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,128,256,0.00281866660548581
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,64,8192,0.006575111299753189
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,64,10240,0.017317333155208163
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,64,7168,0.005910222315125995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,64,6144,0.006275555739800136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,64,5120,0.006260444306664997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,64,8192,0.01633777717749278
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,4,128,256,0.005428444594144821
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,64,4096,0.006102222121424145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,64,7168,0.015508444772826301
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,64,3584,0.00609777785009808
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,64,6144,0.015141333142916361
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,64,3072,0.005927111125654644
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,64,5120,0.014501333236694336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,64,2560,0.005919111271699269
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,64,4096,0.013146667016877068
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,64,2048,0.005936000082227919
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,64,3584,0.013416888813177744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,64,1536,0.0042239998777707415
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,64,3072,0.013134222063753339
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,64,1024,0.003567111161020067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,64,768,0.003489777859714296
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,64,2560,0.01311377767059538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,64,512,0.0031999999450312722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,64,2048,0.011745777395036487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,128,32,0.002495999965402815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,64,1536,0.01239111108912362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,64,256,0.002521777732504739
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,64,1024,0.011907555162906647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,64,128,0.0028124445428450904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,64,64,0.002628444383541743
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,64,768,0.0120319997270902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,64,32,0.002499555548032125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,64,512,0.011735111474990845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,64,256,0.011060444845093621
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,64,128,0.010553778045707278
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,64,64,0.01074577785200543
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,32,16384,0.0064382221963670515
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,32,12288,0.006224888894293044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,64,32,0.01037333326207267
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,32,10240,0.005921777751710679
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,32,8192,0.0055440002017551
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,32,16384,0.02032355633046892
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,32,10240,0.017875555488798354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,32,8192,0.01587377819750044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,32,6144,0.006447111152940326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,32,7168,0.015973332855436537
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,32,5120,0.005929777605666055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,32,6144,0.015104000767072042
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,32,4096,0.005565333283609814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,32,5120,0.014663111832406787
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,32,3584,0.005938666562239329
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,32,4096,0.013747555514176687
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,32,3072,0.005698666804366642
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,32,3584,0.013804444836245643
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,32,2560,0.005935110979610019
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,32,3072,0.013384888569513956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,32,2048,0.00555644432703654
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,32,2560,0.01310400002532535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,32,1536,0.004516444272465176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,32,65536,0.009301332963837517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,32,2048,0.012424888710180918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,32,1536,0.012404444317022959
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,32,768,0.0034933334423436057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,32,65536,0.04860533277193705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,32,1024,0.011762667033407422
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,32,512,0.0031795555518733132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,32,7168,0.005723555468850666
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,32,256,0.0028497777465316984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,32,128,0.0028320000403457214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,32,768,0.01203555530971951
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,64,12288,0.01851288808716668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,32,512,0.011355555719799466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,32,32,0.002512000087234709
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,32,256,0.010026666853162978
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,32,64,0.01073066641887029
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,32,32,0.010097777677906884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,65536,16384,0.19139911068810356
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,65536,16384,0.334187560611301
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,65536,12288,0.2555430995093452
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,32,1024,0.0038408889538711975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,16384,0.13781332969665527
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,65536,12288,0.16224888960520426
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,32,12288,0.018544000056054857
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,65536,10240,0.21456710497538248
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,65536,10240,0.13631644513871935
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,65536,8192,0.17654666635725233
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,4,32,128,0.011733333269755045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,12288,0.10387644502851699
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,10240,0.08852622244093154
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,65536,8192,0.12156889173719619
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,65536,7168,0.14480800098843044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,8192,0.07406044668621488
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,65536,7168,0.10739733113182916
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,65536,6144,0.10611022180981106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,7168,0.06734133428997464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,65536,5120,0.10806044605043198
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,65536,4096,0.0899875561396281
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,6144,0.06145066685146756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,5120,0.055758224593268506
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,65536,3584,0.07232177919811673
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,65536,4096,0.06986222002241346
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,4096,0.05013066530227661
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,65536,3072,0.05936355723275078
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,65536,3584,0.06714577807320489
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,3584,0.04520444406403435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,65536,3072,0.05768444140752157
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,65536,2560,0.05261333452330696
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,4,32,64,0.002512000087234709
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,65536,2048,0.04373777906099955
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,65536,2560,0.04781244529618157
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,3072,0.03977866636382209
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,65536,1536,0.03367200162675645
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,65536,2048,0.039566222164365984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,65536,1024,0.024359110328886244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,2048,0.02941511074701945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,65536,1536,0.03471022182040744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,65536,6144,0.12663288911183676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,1536,0.02367466688156128
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,1024,0.018213333355055917
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,65536,512,0.013527111046844058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,768,0.015216888652907478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,65536,768,0.023369777533743117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,65536,256,0.008077333370844523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,65536,5120,0.08480978012084961
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,65536,512,0.019189332922299702
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,512,0.012718222207493253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,65536,128,0.006487999939256244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,65536,256,0.016524444023768108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,256,0.010467555787828235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,65536,64,0.005705777969625261
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,128,0.010635555618339114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,65536,64,0.013515555196338229
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,65536,32,0.013404444687896304
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,16384,65536,0.3167422347598606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,16384,65536,0.18319288889567056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,65536,2560,0.03558844327926636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,16384,16384,0.08486488792631362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,65536,768,0.01889333294497596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,65536,1024,0.02738488879468706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,16384,16384,0.07689333624309964
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,16384,0.04383733206325107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,16384,12288,0.059862222936418324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,16384,10240,0.05719111363093058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,12288,0.03475733266936408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,65536,0.16316978136698404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,16384,8192,0.04471822248564827
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,16384,10240,0.05407289001676771
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,65536,32,0.005579555614127054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,10240,0.029633777009116277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,16384,7168,0.03868800070550706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,16384,8192,0.04478311207559374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,8192,0.02472977836926778
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,16384,6144,0.033766223324669734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,16384,7168,0.040082666609022356
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,7168,0.023218666513760883
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,16384,5120,0.028748446040683325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,16384,6144,0.03875199953715006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,16384,4096,0.023641778363121882
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,6144,0.0225600004196167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,5120,0.021103999680942957
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,16384,3584,0.020554666717847187
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,16384,4096,0.02784088916248745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,4096,0.018515555395020377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,16384,3072,0.018569777409235638
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,3584,0.018360000517633226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,16384,3584,0.026383111874262493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,16384,2560,0.015440000428093804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,3072,0.01683555543422699
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,16384,2048,0.013447999954223633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,16384,12288,0.06517777840296428
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,16384,2560,0.021592888567182753
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,16384,1536,0.010990222295125326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,2560,0.015124445160230001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,16384,2048,0.019653333557976615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,2048,0.013081777426931592
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,1536,0.011052444577217102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,16384,768,0.0069004446268081665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,16384,1024,0.016140444411171805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,1024,0.009217778013812171
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,65536,128,0.014091556270917257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,16384,512,0.005201777650250329
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,16384,768,0.01346933344999949
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,768,0.008173333273993598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,16384,256,0.003495111233658261
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,16384,512,0.01311288856797748
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,512,0.007278222176763747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,16384,128,0.0032008888406885993
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,16384,256,0.012715555727481842
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,256,0.006592000110281839
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,16384,64,0.002814222127199173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,16384,5120,0.031664000617133245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,16384,128,0.010695111420419482
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,16384,128,0.006619555668698416
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,16384,64,0.010694222317801582
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,16384,32,0.010030222435792288
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,16384,3072,0.024373332659403484
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,12288,65536,0.2397866778903537
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,12288,16384,0.06749511427349515
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,12288,65536,0.1523537768257989
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,16384,1024,0.008430221842394935
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,16384,1536,0.01828266680240631
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,12288,12288,0.05120088987880283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,12288,16384,0.06351821952395968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,16384,0.038650665018293597
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,65536,0.14516711235046387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,12288,10240,0.043712000052134194
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,12288,12288,0.05320177806748284
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,12288,8192,0.03798933492766486
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,12288,0.03050400151146783
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,12288,7168,0.03429155548413595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,10240,0.026086222794320848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,12288,8192,0.039552887280782066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,12288,6144,0.03018488817744785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,12288,7168,0.035340444909201726
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,7168,0.02089777754412757
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,12288,5120,0.02605244517326355
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,12288,6144,0.03463644451565213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,6144,0.01923644377125634
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,12288,4096,0.02272355556488037
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,12288,5120,0.027592889136738245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,16384,32,0.0028702221396896574
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,5120,0.018545778261290658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,12288,3584,0.019091554813914828
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,12288,4096,0.026769777139027912
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,4096,0.01704088846842448
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,12288,3072,0.017504000001483493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,12288,2560,0.015138667490747241
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,3584,0.015768888923856948
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,12288,3584,0.024999999337726172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,12288,2048,0.013766222529941134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,12288,3072,0.02307644486427307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,12288,2560,0.020006222857369315
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,2560,0.01294222225745519
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,12288,1536,0.008652444514963362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,2048,0.011090666883521609
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,12288,2048,0.018505778577592637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,12288,1024,0.0069520001610120135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,12288,1536,0.017367111312018502
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,1536,0.010138666464222802
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,12288,768,0.006050666587220297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,12288,1024,0.014434667097197639
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,1024,0.00829777783817715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,12288,512,0.0038133333954546186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,12288,768,0.01309333327743742
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,12288,256,0.0034799998005231223
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,12288,512,0.012768000364303589
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,12288,10240,0.04607111215591431
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,12288,128,0.0028168888141711554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,8192,0.022091555926534865
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,12288,256,0.011729777687125735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,256,0.006581333362393909
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,12288,64,0.0028275555620590844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,12288,32,0.002847111059559716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,128,0.006362666686375936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,12288,128,0.011406222151385413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,12288,64,0.010947555303573608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,12288,32,0.009689778089523315
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,10240,65536,0.20467200544145372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,10240,16384,0.05729689200719198
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,10240,65536,0.15007377995385066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,10240,12288,0.04433066646258036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,3072,0.015019555886586508
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,16384,0.04270044300291273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,65536,0.1351591083738539
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,10240,16384,0.056530667675866016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,10240,10240,0.0382008883688185
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,12288,0.0328764451874627
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,10240,8192,0.031897776656680636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,10240,12288,0.044687112172444664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,10240,0.025360888904995386
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,10240,7168,0.02927288744184706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,8192,0.021451556020312842
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,10240,8192,0.03425066669782003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,768,0.007598222129874759
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,10240,7168,0.03035466538535224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,10240,6144,0.025792888469166223
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,7168,0.01992266707950168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,12288,512,0.0069448889957533936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,10240,4096,0.01959822244114346
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,6144,0.018907555275493197
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,5120,0.01759822169939677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,10240,5120,0.02440799938307868
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,10240,3584,0.01717511150572035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,10240,4096,0.023335110810067918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,4096,0.015493333339691162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,10240,3072,0.01683466633160909
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,3584,0.015232889188660515
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,10240,3072,0.020389333367347717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,3072,0.013810666898886362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,10240,2560,0.015435554915004306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,2560,0.011752888560295105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,10240,2048,0.010880000061459012
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,10240,10240,0.04144533475240072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,10240,2048,0.017900443739361234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,10240,1536,0.009164444274372524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,2048,0.010293333066834344
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,10240,1024,0.006559999866618051
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,10240,5120,0.022713777091768052
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,10240,1536,0.016418667303191293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,1536,0.009308444129096137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,10240,768,0.004902222090297275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,10240,6144,0.02868088748719957
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,10240,512,0.003479111111826367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,10240,1024,0.013795555465751223
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,1024,0.00796888851457172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,10240,256,0.0031555555760860443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,10240,768,0.013452444639470843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,768,0.0074444446298811175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,10240,128,0.002883555574549569
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,512,0.0069120000633928515
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,10240,512,0.01274222218328052
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,10240,64,0.002785777880085839
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,256,0.00630044440428416
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,10240,256,0.012055111428101858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,10240,32,0.0028515555378463534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,10240,128,0.0063493334584765965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,10240,64,0.01036977767944336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,10240,32,0.01035199976629681
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,8192,65536,0.16862755351596406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,8192,16384,0.04661155409283108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,8192,65536,0.09786222378412883
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,10240,2560,0.019150222341219585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,8192,12288,0.03655111127429538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,16384,0.02351911034848955
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,8192,16384,0.04359466830889384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,65536,0.07799822092056274
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,8192,10240,0.029496000872717962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,8192,12288,0.035522666242387556
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,12288,0.019289778338538278
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,8192,8192,0.024545777175161574
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,8192,7168,0.02200177808602651
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,8192,10240,0.031151112582948472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,8192,8192,0.02757688860098521
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,8192,0.015035554766654968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,8192,6144,0.01917155583699544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,8192,7168,0.025068443682458665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,7168,0.014039110806253223
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,8192,5120,0.01644888851377699
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,8192,6144,0.024700444605615404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,6144,0.013797333670987023
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,8192,4096,0.014440000057220459
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,8192,5120,0.021021333005693223
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,5120,0.01238044434123569
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,8192,3584,0.012698666916953193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,8192,4096,0.02032533288002014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,4096,0.011715555356608497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,10240,3584,0.02164266672399309
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,10240,128,0.011338666909270816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,8192,3584,0.01887911061445872
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,3584,0.010900444454616971
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,8192,2560,0.009900444083743626
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,8192,2048,0.008604444563388824
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,3072,0.00999111102686988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,2560,0.009018667042255402
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,8192,1536,0.007247111035717859
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,8192,2048,0.015662221444977652
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,2048,0.00831733312871721
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,8192,1024,0.005250666704442766
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,8192,1536,0.01408622165520986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,1536,0.007601777712504069
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,8192,768,0.004176889028814104
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,8192,1024,0.0120666664507654
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,1024,0.006631999793979857
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,8192,512,0.003559999995761447
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,10240,0.01759199963675605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,8192,768,0.012755555411179861
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,768,0.0064426664676931165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,8192,256,0.0032088889016045462
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,8192,512,0.011713777979214987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,512,0.006286222073766921
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,8192,128,0.003154666680428717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,8192,256,0.011085333095656501
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,256,0.005897777775923411
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,8192,64,0.0028417776856157514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,8192,32,0.0028364445186323593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,8192,128,0.01072355525361167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,8192,128,0.00592533333433999
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,8192,64,0.01037333326207267
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,8192,32,0.009395555489593083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,8192,3072,0.011163555913501315
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,7168,65536,0.1461226675245497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,7168,16384,0.04199022385809156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,7168,65536,0.09090489149093628
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,8192,3072,0.018515555395020377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,7168,12288,0.03372711274358962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,8192,2560,0.016582222448454965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,16384,0.02256444427702162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,7168,16384,0.040729777680502996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,7168,10240,0.029334220621320937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,7168,12288,0.03484266665246751
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,7168,8192,0.025640888346566096
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,12288,0.01852888862291972
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,10240,0.01650311052799225
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,7168,10240,0.029491557015313044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,7168,8192,0.027845333019892376
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,8192,0.014135110709402295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,7168,6144,0.018530666828155518
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,7168,7168,0.024734222226672705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,7168,5120,0.014474666780895658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,7168,6144,0.022897778285874262
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,6144,0.012379555238617791
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,7168,4096,0.013776889277829064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,7168,5120,0.021215111017227173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,5120,0.011401777466138205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,7168,4096,0.01958488921324412
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,7168,3584,0.018423111902342904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,7168,3072,0.01070577816830741
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,7168,3072,0.017849778135617573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,65536,0.07384444607628717
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,3072,0.008627555436558193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,7168,2560,0.008934222161769867
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,7168,7168,0.01920088960064782
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,7168,2560,0.015801777442296345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,2560,0.008284444610277811
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,7168,2048,0.008095111283991072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,7168,1536,0.00657155571712388
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,7168,0.013268444273206921
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,7168,2048,0.014496000276671516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,2048,0.007599111232492659
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,7168,1024,0.004874666531880696
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,7168,1536,0.013765333427323235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,1536,0.0069511110583941145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,7168,768,0.003834666477309333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,1024,0.006299555715587404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,7168,512,0.0035866668654812705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,7168,768,0.011739555332395764
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,768,0.006248000181383557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,4096,0.010006222460005019
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,7168,256,0.0031537777847713898
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,7168,512,0.011715555356608497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,512,0.005953777581453323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,7168,128,0.0028799999919202593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,3584,0.009344889058007134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,7168,256,0.011728888584507836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,256,0.005643555687533484
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,7168,64,0.0028124445428450904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,7168,32,0.0028115556471877625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,7168,128,0.005630222045713001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,7168,128,0.010080888867378235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,7168,64,0.01035822182893753
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,7168,32,0.010336000058386061
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,6144,65536,0.1252382198969523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,6144,16384,0.037159999211629234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,6144,65536,0.08383999930487739
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,6144,12288,0.029172443681293066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,6144,16384,0.03836089041497972
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,65536,0.07130399677488539
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,16384,0.02204533252451155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,12288,0.017856889300876193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,6144,8192,0.021344888541433547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,10240,0.01589777734544542
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,8192,0.013952889376216464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,6144,8192,0.02531911101606157
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,6144,7168,0.019883554842736986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,7168,1024,0.012727999852763282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,7168,3584,0.011962666279739805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,6144,7168,0.023363555471102398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,7168,0.013036444783210754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,6144,5120,0.015847111741701763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,6144,6144,0.02292177743381924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,6144,0.012147555748621622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,6144,4096,0.011705777711338468
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,5120,0.011027555498811932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,6144,5120,0.01958755486541324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,6144,3584,0.010663111176755695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,6144,4096,0.018619555566045973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,4096,0.009336888790130615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,6144,3072,0.00964355551534229
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,6144,3584,0.01812000075976054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,3584,0.009287111461162567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,6144,2560,0.00832622249921163
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,6144,10240,0.025618667403856914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,6144,3072,0.01754666699303521
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,3072,0.008295999632941352
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,6144,2048,0.007605333295133378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,6144,10240,0.027936001618703205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,2560,0.00793599999613232
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,6144,1536,0.006272888845867581
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,6144,2048,0.014053333136770459
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,2048,0.007594666547245449
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,6144,1024,0.004218666503826777
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,1536,0.006665777828958299
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,6144,1536,0.013799111048380533
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,6144,768,0.003839111162556542
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,6144,1024,0.012724444270133972
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,6144,6144,0.017999110950364005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,1024,0.006268444574541516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,6144,512,0.003169777699642711
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,6144,768,0.012437333663304647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,768,0.006248000181383557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,6144,256,0.002867555452717675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,6144,512,0.010778666370444827
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,6144,128,0.0028417776856157514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,256,0.005883555445406172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,6144,64,0.002480000050531493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,6144,128,0.010378667049937779
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,128,0.005892444401979446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,6144,32,0.0028373334142896864
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,6144,64,0.009700444837411245
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,6144,32,0.010159999959998662
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,5120,65536,0.10863910781012641
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,5120,16384,0.032933334509531655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,6144,12288,0.031247112486097548
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,5120,65536,0.07540533277723524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,6144,2560,0.015388444066047668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,5120,12288,0.026947556270493403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,5120,16384,0.03628888726234436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,16384,0.02111022174358368
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,5120,10240,0.02293599976433648
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,12288,0.017504889104101393
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,5120,12288,0.03037244412634108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,5120,10240,0.027789334456125896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,10240,0.015463999576038785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,5120,7168,0.017857778403494094
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,5120,8192,0.024708444873491924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,8192,0.013576889203654395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,5120,6144,0.01616533266173469
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,6144,512,0.005944889038801193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,5120,7168,0.02276533345381419
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,7168,0.012709333664841123
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,6144,256,0.011749332977665795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,5120,5120,0.014319111903508505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,5120,4096,0.010686222049925061
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,5120,6144,0.02126400007141961
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,5120,0.010055999788973067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,5120,5120,0.019499555230140686
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,5120,3584,0.009644444617960189
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,4096,0.008980444735950893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,5120,4096,0.017885333961910672
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,5120,3072,0.008719111482302347
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,3584,0.008282666405042013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,5120,2560,0.00776088900036282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,5120,3072,0.016181333197487723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,5120,2560,0.014554666148291694
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,2560,0.007283555964628856
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,65536,0.06932266553243001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,5120,2048,0.00685777763525645
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,5120,1536,0.005220444252093633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,2048,0.007257777783605788
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,5120,2048,0.01348888874053955
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,5120,1024,0.003920888735188378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,5120,1536,0.013399110900031196
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,1536,0.0069022224181228215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,5120,768,0.0038453332251972626
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,5120,1024,0.011390222443474663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,1024,0.006299555715587404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,5120,512,0.0035013332962989807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,5120,768,0.01240711079703437
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,768,0.0059164443777667145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,5120,256,0.0028453332682450614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,5120,512,0.011696000066068439
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,6144,0.011688888900809817
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,512,0.005943111247486538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,5120,128,0.0027786667148272195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,5120,256,0.010993777877754636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,256,0.005720888988839255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,5120,64,0.0025031111306614345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,128,0.0058986664646201665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,5120,32,0.0028364445186323593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,5120,128,0.011076444553004371
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,5120,3584,0.017480888300471835
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,5120,64,0.010348444183667501
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,4096,65536,0.09071644147237141
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,5120,32,0.010041777458455827
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,5120,3072,0.007877333296669854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,4096,12288,0.02204799983236525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,4096,65536,0.06826488839255439
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,4096,10240,0.01885511146651374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,16384,0.02053511142730713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,4096,12288,0.02739022175470988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,65536,0.06689244508743286
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,12288,0.016754666964213055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,4096,8192,0.01629599928855896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,4096,7168,0.01480977733929952
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,4096,10240,0.02609066665172577
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,4096,8192,0.023186667097939387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,8192,0.013097777962684631
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,4096,6144,0.01349066694577535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,7168,0.012058667010731168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,4096,7168,0.02148177723089854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,4096,5120,0.011867555479208628
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,4096,6144,0.021132444341977436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,6144,0.010716444088353051
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,4096,4096,0.010760889285140567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,4096,5120,0.018955555227067735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,5120,0.009972444838947719
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,4096,3584,0.011397333608733283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,4096,4096,0.01735822194152408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,4096,0.008340444829728868
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,4096,3072,0.010086221827401055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,4096,3584,0.016467556357383728
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,3584,0.007991111112965478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,4096,2560,0.008623999853928884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,4096,3072,0.015102222561836243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,4096,2560,0.013449778159459433
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,4096,2048,0.007249777515729268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,4096,16384,0.032820443312327065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,4096,2048,0.013408888545301227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,4096,1536,0.005922666854328579
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,10240,0.015102222561836243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,2048,0.0069511110583941145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,4096,1024,0.005573333137565189
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,5120,8192,0.0199644449684355
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,4096,1536,0.013059555656380124
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,1536,0.006833777659469181
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,4096,768,0.004491555607981152
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,4096,1024,0.011384888655609555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,1024,0.006235555642180973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,4096,512,0.0041982221106688184
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,768,0.006231110956933763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,4096,768,0.012429333395428128
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,4096,256,0.0031439999325407874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,4096,512,0.011724444727102915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,4096,256,0.011764444410800934
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,256,0.005589333259397083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,128,0.005910222315125995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,4096,128,0.01036800030204985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,4096,64,0.002522666628162066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,4096,32,0.002797333316670524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,3072,0.007941332956155142
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,4096,16384,0.02834577692879571
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3584,65536,0.07936444547441271
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,4096,32,0.010031110710567897
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3584,16384,0.02495288848876953
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3584,65536,0.06506133079528809
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3584,16384,0.03012977706061469
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,65536,0.0661075578795539
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3584,12288,0.019471999671724107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,16384,0.020278222031063504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3584,10240,0.017163554827372234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3584,12288,0.027504889501465693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,12288,0.016557332542207506
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3584,8192,0.01496088835928175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3584,10240,0.024439111351966858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,10240,0.014735110931926303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3584,7168,0.014562666416168213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3584,8192,0.02164177762137519
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,8192,0.012708444562223224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,4096,128,0.0028035555862718155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3584,6144,0.012425777812798819
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3584,7168,0.02127733329931895
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,7168,0.011405333048767514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3584,5120,0.012135110795497894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3584,6144,0.020219556159443326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,6144,0.010061333576838175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3584,4096,0.010650667051474253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3584,5120,0.017500445246696472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,5120,0.009333333207501305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3584,3584,0.01131199962562985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,4096,64,0.00963288876745436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3584,4096,0.015572445260153877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,4096,0.008328888979223039
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,2560,0.007596444752481248
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3584,2560,0.008620444271299574
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,3584,0.00797155582242542
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3584,3072,0.014089778065681458
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3584,2048,0.007298666569921706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,2560,0.007599111232492659
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3584,2560,0.013800889253616333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3584,1536,0.005928888916969299
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3584,2048,0.012919999659061432
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,2048,0.007250666618347168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3584,1024,0.004864000197913912
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3584,1536,0.013411555853154926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,1536,0.0069511110583941145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3584,768,0.004182222402758068
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3584,1024,0.01202933324707879
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,1024,0.00628355559375551
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3584,512,0.0038773334688610504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,4096,512,0.00591466658645206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,768,0.005959110955397288
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3584,256,0.0031306667046414483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3584,512,0.011741333537631564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,512,0.0059057776298787855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3584,128,0.0028488888508743713
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3584,256,0.011686222420798408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3584,128,0.010066666536860997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,128,0.005879999862776862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3584,64,0.0025537777692079544
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3584,32,0.0028231110837724474
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3584,3072,0.01075466639465756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3584,3584,0.015848888291252982
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3584,64,0.010336000058386061
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,3072,0.007974222302436829
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3584,32,0.009346666435400644
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3072,16384,0.021518222159809534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3072,65536,0.06274755795796712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3072,12288,0.01799199978510539
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,65536,0.06527822547488742
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,16384,0.019967110620604623
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3072,16384,0.02978666623433431
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3072,10240,0.01591822173860338
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3072,12288,0.025861332813898723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3072,10240,0.023471110396915015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,10240,0.01460355520248413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3584,768,0.01237511138121287
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3072,7168,0.012713777522246042
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3072,8192,0.021031111478805542
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,7168,0.0107342220014996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3072,7168,0.019941333267423842
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3072,6144,0.011359111302428775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3072,6144,0.018732445107565988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,6144,0.01000355515215132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3072,5120,0.010784000158309937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,5120,0.009320889082219865
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3072,4096,0.009549332989586724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3072,4096,0.01516266663869222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,12288,0.01647822227742937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3072,3584,0.00922400007645289
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3072,3584,0.014468444718254937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,8192,0.012154666913880242
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3072,3072,0.00927022182279163
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,3584,0.007997333175606197
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,3584,256,0.00590844452381134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3072,2560,0.007575111256705389
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3072,3072,0.014152889450391134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,3072,0.007923555870850882
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3072,2048,0.006559111177921295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3072,2560,0.013400889105266996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,2560,0.007624000310897827
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3072,1536,0.006265777680608962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3072,2048,0.013088000317414602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3072,5120,0.016470222009552848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3072,1024,0.005199999858935674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3072,1536,0.012440889245933957
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,1536,0.006919999917348226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3072,768,0.004222222086456087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,1024,0.0064311110311084324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3072,768,0.011753777662913004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,768,0.006228444476922353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3072,512,0.0035439998739295532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3072,8192,0.013770666387346057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3072,256,0.003491555651028951
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,512,0.0059164443777667145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3072,512,0.011723555624485016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3072,128,0.0032088889016045462
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3072,256,0.011694221860832639
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,256,0.005767999837795894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3072,64,0.002512000087234709
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3072,128,0.010386666489972008
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,128,0.005882666756709416
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3072,32,0.0028613333900769553
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3072,64,0.010029333333174387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2560,65536,0.0594773358768887
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3072,32,0.010064000056849586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2560,16384,0.019550222489568923
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2560,65536,0.06035377581914266
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,2048,0.007264888948864407
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,65536,0.0633466641108195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,3072,65536,0.06975199778874715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2560,12288,0.017058667209413316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2560,16384,0.026426666312747534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2560,10240,0.013895110951529609
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2560,12288,0.0243848893377516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,12288,0.015784889459609985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,3072,1024,0.012054222325483957
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2560,8192,0.012724444270133972
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,10240,0.013986666997273764
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2560,7168,0.01202311118443807
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,8192,0.011422221859296164
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,3072,4096,0.00830577810605367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2560,6144,0.010703999963071613
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2560,6144,0.017316444052590262
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,6144,0.010034666293197209
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2560,5120,0.009931555224789513
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2560,5120,0.014775110615624322
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,5120,0.009328000247478485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2560,4096,0.009113777842786577
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2560,3584,0.008650666309727563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,4096,0.008283555507659912
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2560,3584,0.0141004439857271
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2560,3072,0.007824888659848107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,3584,0.00793866647614373
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2560,3072,0.013432000246312885
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,3072,0.0076453329788313965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2560,2560,0.006592000110281839
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2560,10240,0.021612443857722815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,2560,0.007292444507280986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2560,2048,0.006257777826653586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2560,8192,0.020222221811612446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,2048,0.006974221931563483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2560,2048,0.012424888710180918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2560,7168,0.018915555543369718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2560,1536,0.005864000154866113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2560,1024,0.00518844442235099
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,1536,0.006928000185224745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2560,1536,0.012075555821259817
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2560,768,0.0041991112132867174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2560,1024,0.011761777930789523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,1024,0.006335111127959357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2560,512,0.0038897775941424896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2560,768,0.011056000159846412
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2560,512,0.01146044416560067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2560,4096,0.014101333088344999
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2560,256,0.003167111012670729
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,256,0.005597333527273602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2560,128,0.003145777723855443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,16384,0.019711110326978896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2560,128,0.011365333365069495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2560,64,0.002528000002106031
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2560,2560,0.013086222112178802
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2560,32,0.002630222174856398
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2560,64,0.010381333529949188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,7168,0.01071200023094813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2560,32,0.009992000129487779
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2048,65536,0.051206221183141075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2048,16384,0.01647111111217075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2048,65536,0.058423108524746366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,65536,0.05977244509590996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2048,12288,0.013574221895800697
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2048,16384,0.025378665990299646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,768,0.005929777605666055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,16384,0.019343111250135634
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2048,10240,0.012088889049159156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,12288,0.015088889333936902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2048,12288,0.02398666739463806
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2048,8192,0.011001778145631155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,512,0.00590133335855272
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2048,10240,0.020871111088328894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2048,7168,0.010406222608354356
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,10240,0.013131555583741931
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2048,8192,0.01995822290579478
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2048,6144,0.009397333694828881
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2560,256,0.011720000041855706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,7168,0.01037333326207267
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2048,7168,0.017318222257826064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,6144,0.009995555712117089
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2048,4096,0.010308444499969482
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,5120,0.009369778136412302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2048,5120,0.014451555079884
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2048,3584,0.009405333134863112
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,2560,128,0.005599110904667113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2048,4096,0.014122666584120857
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2048,3072,0.006631111105283101
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2048,3584,0.013743110828929476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2048,3072,0.013608888619475894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,3072,0.007669332954618666
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2048,2560,0.006232000059551663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2048,2560,0.01255644443962309
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,2560,0.007304888632562425
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2048,2048,0.007270221908887227
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2048,2048,0.012715555727481842
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,2048,0.006950221955776215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2048,1536,0.005600000007285013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2048,1536,0.012111110819710625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,8192,0.011365333365069495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2048,1024,0.004544888933499654
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2048,5120,0.00870666652917862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2048,768,0.004583999928500918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2048,6144,0.01615111033121745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2048,1024,0.011742221812407175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,1024,0.006288888967699475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2048,512,0.0038515557017591265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2048,768,0.011933333343929715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,768,0.006244444598754247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2048,256,0.0034133332471052804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,512,0.005881777654091517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,4096,0.008533333738644918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2048,128,0.0031395554542541504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,3584,0.007995555798212687
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,256,0.00556355549229516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2048,64,0.002809777855873108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,128,0.005594666633341048
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2048,128,0.01035911093155543
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1536,65536,0.037717332442601524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2048,64,0.010026666853162978
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2048,32,0.010045333041085137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1536,16384,0.013738666971524557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1536,65536,0.05670577949947781
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1536,16384,0.024716445141368445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1536,12288,0.011551111108726926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,16384,0.01739911072784
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,2048,1536,0.006618666566080517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1536,12288,0.021645334031846788
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,12288,0.014129777749379476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1536,10240,0.019315555691719055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,10240,0.012797333300113678
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1536,8192,0.01036622209681405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2048,512,0.010399111443095738
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1536,8192,0.016945777667893302
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1536,7168,0.009729777773221334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,2048,256,0.011713777979214987
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,2048,32,0.0027991111079851785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1536,6144,0.009001777403884465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1536,7168,0.015146666102939181
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1536,5120,0.007632888853549957
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,7168,0.0103502223889033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,6144,0.00977955593003167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1536,4096,0.007582222421964009
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,5120,0.009036444127559662
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1536,5120,0.014108444253603617
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1536,3584,0.0069360000391801195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,65536,0.05883644686804878
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,4096,0.008289777570300633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1536,3072,0.006588444527652528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1536,3584,0.012862222062216865
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,3584,0.007932444413503012
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1536,2560,0.005642666584915585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,3072,0.007738666401969061
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1536,3072,0.013429332938459186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1536,10240,0.010653333531485664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1536,2048,0.006244444598754247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1536,2560,0.012784889174832238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1536,1536,0.005240889059172736
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,2560,0.007282666862010956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1536,2048,0.012381333443853589
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1536,1024,0.004184000194072723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1536,1536,0.012216889195972018
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,1536,0.00665866666369968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1536,768,0.0038622220357259116
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,8192,0.011039999624093374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1536,1024,0.011736888852384357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,1024,0.006275555739800136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1536,512,0.003504000190231535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1536,768,0.01074044406414032
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,768,0.006215111249023014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1536,256,0.0028444443725877335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,512,0.0058906666106647914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1536,128,0.002797333316670524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1536,256,0.012068444656001197
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,256,0.005597333527273602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1536,6144,0.015505777464972602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1536,64,0.0026746667507621977
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1536,32,0.0028391112056043413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1536,128,0.010991111397743225
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,128,0.005585777676767773
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1536,64,0.00998488896422916
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1536,4096,0.014095111025704278
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1536,32,0.009370666411187913
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1024,16384,0.014535110857751636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1024,65536,0.05442844496832954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,65536,0.058563556936052114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1024,16384,0.022988445229000513
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,16384,0.016880000631014507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1024,10240,0.011432888607184092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1024,12288,0.018792000081804063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,12288,0.013752000199423896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1024,8192,0.010211555494202508
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,1536,2048,0.0069448889957533936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1024,10240,0.01754577789041731
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,10240,0.01275466630856196
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1024,7168,0.009752000371615091
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,8192,0.011065777805116443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1024,8192,0.016187555260128446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1024,6144,0.008504889077610439
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,7168,0.010401777923107147
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1024,7168,0.015835555063353646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1024,5120,0.007941332956155142
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1024,6144,0.01478844384352366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,6144,0.009672000176376766
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1024,4096,0.007164444360468123
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1024,5120,0.014470222923490735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1536,512,0.011008888483047485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,5120,0.009175111022260454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1024,3584,0.006913777854707506
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1024,3072,0.006264888991912206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,4096,0.007976888782448238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,3584,0.007930667036109501
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1024,2560,0.005591111050711737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1024,3072,0.013462222284740873
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,3072,0.007620444728268518
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1024,65536,0.026222222381167944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1024,2048,0.006577777779764599
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1024,2560,0.012774222426944308
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1024,1536,0.005533333453867171
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1024,2048,0.01237422227859497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,2048,0.0069368887278768755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1024,12288,0.012429333395428128
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1024,1024,0.005567111074924469
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,1536,0.006671999891599019
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,1024,0.006270222365856171
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1024,1024,0.01168444421556261
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,768,0.005973333285914526
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1024,512,0.0038497779104444715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1024,512,0.011416888899273343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,512,0.005896889087226655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1024,256,0.0032311110860771606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1024,4096,0.013778666655222574
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1024,3584,0.013398222625255585
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1024,128,0.0028684443483750024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1024,256,0.012071111136012606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,256,0.005616000129116907
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,2560,0.007291555404663086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1024,128,0.011063111325105032
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1024,64,0.002801777794957161
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1024,32,0.0028159999185138275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1024,64,0.010326222413116032
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1024,1536,0.012327111429638334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1024,32,0.009687999884287516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,1024,768,0.004243555582231945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,768,65536,0.021706667211320665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,768,16384,0.01035822182893753
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,768,12288,0.010861333045694562
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,768,65536,0.052720887793434985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,1024,768,0.01070755554570092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,65536,0.05843466520309448
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,768,16384,0.02144711050722334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,16384,0.01683555543422699
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,768,10240,0.009677333136399588
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,768,8192,0.009008888569143083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,768,12288,0.0185244447655148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,12288,0.014104000396198697
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,768,7168,0.008620444271299574
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,768,10240,0.01757955551147461
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,768,8192,0.015990222493807476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,768,6144,0.007993777592976889
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,8192,0.011167111496130625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,768,5120,0.0069475554757648045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,7168,0.010665777656767102
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,768,7168,0.015799999237060547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,768,4096,0.007250666618347168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,768,6144,0.014801777071423002
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,5120,0.009191110730171204
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,768,3584,0.006937777830494775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,4096,0.00797333319981893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,768,3584,0.013447110851605734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,3584,0.007735110819339752
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,768,3072,0.006566222343179915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,1024,128,0.005575110928879843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,768,2560,0.006320888797442119
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,3072,0.007631999750932057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,768,2560,0.012721777790122561
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,2560,0.007304888632562425
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,768,2048,0.0052266667286554975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,2048,0.0069822221994400024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,768,2048,0.012383999923865
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,768,1536,0.005208000126812193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,10240,0.012531555361217923
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,1536,0.006635555376609166
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,768,1024,0.004243555582231945
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,1024,0.006275555739800136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,6144,0.009992000129487779
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,768,768,0.003896000070704354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,768,5120,0.014648000399271647
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,768,512,0.003528888854715559
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,768,4096,0.014072888427310519
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,768,0.005960000058015187
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,768,256,0.003168888803985384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,768,512,0.011383111278216044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,512,0.00582755563987626
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,768,128,0.0028577778074476454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,768,256,0.012039110892348819
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,256,0.005600000007285013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,768,3072,0.012788444757461548
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,768,64,0.002491555487116178
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,768,128,0.011009777585665384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,768,32,0.0024826667375034755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,512,65536,0.01719377769364251
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,768,64,0.009352000223265754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,768,32,0.010037333601050908
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,512,16384,0.01070755554570092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,512,65536,0.05233777893914116
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,65536,0.05167555477884081
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,512,12288,0.008660444782839881
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,512,16384,0.021262221866183813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,16384,0.01684888866212633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,768,1536,0.012328888807031842
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,512,10240,0.007648000286685095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,512,12288,0.018227555685573153
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,512,8192,0.007974222302436829
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,12288,0.01408088869518704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,512,10240,0.017504889104101393
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,10240,0.012727110750145383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,512,7168,0.007620444728268518
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,8192,0.011062222222487131
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,512,8192,0.015818667080667283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,512,6144,0.007251555720965068
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,768,768,0.011741333537631564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,512,5120,0.0069413334131240845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,512,6144,0.014720888601409065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,6144,0.009755555954244401
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,512,4096,0.007256888680987888
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,512,5120,0.014494222071435718
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,5120,0.009001777403884465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,512,3584,0.006763555523422029
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,512,4096,0.01390044391155243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,768,128,0.005561777700980504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,4096,0.008434666527642144
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,512,3072,0.006590222318967183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,512,3584,0.01276533305644989
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,3584,0.007988444632954067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,512,3072,0.013400889105266996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,3072,0.007689778175618913
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,2560,0.007311111523045435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,512,2048,0.005924444645643234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,2048,0.007024889190991719
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,512,1536,0.00452622233165635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,768,1024,0.011755555868148804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,1536,0.00664444433318244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,512,1536,0.012423111332787408
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,512,7168,0.01591911084122128
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,512,1024,0.0038524443904558816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,512,768,0.0034933334423436057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,1024,0.006257777826653586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,512,768,0.011767999993430244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,512,512,0.003175111073586676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,768,0.00591555568906996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,512,512,0.011072888970375061
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,512,0.005923555543025334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,512,256,0.0028275555620590844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,512,2560,0.00628355559375551
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,512,128,0.0028515555378463534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,256,0.005616000129116907
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,512,256,0.011707555916574268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,512,64,0.002502222235004107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,512,2560,0.01298577752378252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,512,32,0.0028133332315418455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,512,64,0.009678222239017487
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,256,65536,0.015842666228612263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,512,2048,0.01272266689274046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,512,32,0.00981866651111179
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,256,16384,0.009007111191749573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,65536,0.05041866501172384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,256,16384,0.021202666891945735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,256,12288,0.008277333445019191
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,16384,0.016862221890025668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,512,1024,0.011705777711338468
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,256,10240,0.007637333538797166
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,256,12288,0.01889511115021176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,12288,0.013801777528391944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,256,8192,0.007246221933099959
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,256,10240,0.01753866672515869
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,10240,0.012735111018021902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,256,7168,0.006624889042642381
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,256,8192,0.016538666354285348
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,8192,0.01106844428512785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,256,6144,0.006276444428496891
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,7168,0.010482666393121084
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,256,6144,0.015497777197096082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,128,0.005607111172543631
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,256,5120,0.005931555396980709
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,256,5120,0.014152000347773233
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,5120,0.00905777762333552
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,256,4096,0.005895111295912001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,256,65536,0.04903199937608507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,4096,0.007999111380841997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,256,3584,0.006261333409282897
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,512,7168,0.010693333215183683
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,256,3072,0.006234666539563074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,3584,0.007967111137178209
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,256,3072,0.013401777380042605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,256,2560,0.006301333506902059
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,3072,0.007662222617202335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,256,2560,0.013072888884279462
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,256,2048,0.005933333188295364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,256,7168,0.015466666883892484
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,256,2048,0.01238044434123569
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,256,1536,0.0042408886882993905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,512,128,0.011047999891969891
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,256,1536,0.012372444073359171
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,1536,0.006620444357395172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,256,1024,0.0035173334181308746
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,1024,0.006311999840868845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,256,1024,0.011413333316644033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,256,768,0.0034924443397257063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,256,4096,0.013798221945762634
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,256,512,0.0031511110977994073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,768,0.005936000082227919
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,256,3584,0.013422222601042854
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,256,512,0.011369778050316704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,512,0.0058711109062035876
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,256,128,0.0028035555862718155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,256,256,0.011709333293967776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,256,0.005607111172543631
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,256,64,0.0028080000645584534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,256,128,0.010997333460383944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,128,0.005618666609128316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,256,32,0.002469333302643564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,256,64,0.010370666782061258
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,128,65536,0.014498665928840637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,256,32,0.0099973330895106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,128,16384,0.00795555528667238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,65536,0.04996888836224874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,128,65536,0.04862044586075676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,2048,0.0069360000391801195
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,128,12288,0.007302222152551015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,128,16384,0.020652444826232064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,16384,0.01680266691578759
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,128,10240,0.00692622239391009
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,12288,0.013799111048380533
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,128,12288,0.018577777677112155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,128,8192,0.006983111302057902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,10240,0.012453333371215396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,128,8192,0.016522667474216886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,8192,0.011026666396194033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,128,7168,0.006602666858169768
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,256,768,0.012048000262843238
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,128,6144,0.00656088896923595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,7168,0.010375111467308467
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,128,7168,0.015474667151769003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,128,5120,0.005903111149867375
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,128,6144,0.015271110667122735
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,128,5120,0.014209777116775513
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,2560,0.007304000357786815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,128,4096,0.005793777604897817
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,5120,0.009012444151772393
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,128,3584,0.006228444476922353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,4096,0.007965332931942409
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,128,3584,0.013415111435784234
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,128,3072,0.006223111102978389
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,3584,0.007967111137178209
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,3072,0.0076044441925154785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,128,3072,0.013187555803192986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,128,2560,0.006253333141406377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,128,10240,0.017239999439981248
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,128,2048,0.005922666854328579
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,2560,0.007358222372002072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,2048,0.0069324444565508105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,128,2048,0.012355555262830524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,256,256,0.003121777748068174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,128,1536,0.01236977759334776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,1536,0.006611555400821898
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,128,1024,0.0038240001433425476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,1024,0.006274666637182236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,128,768,0.0034977777136696708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,128,4096,0.013784888717863293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,128,768,0.01205244412024816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,128,512,0.0031893334041039148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,256,6144,0.009806222385830348
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,128,512,0.011077333655622272
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,512,0.005881777654091517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,128,256,0.0028684443483750024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,256,0.005582222094138463
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,128,256,0.011417778001891242
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,128,128,0.0027875554644399216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,128,128,0.010685332947307162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,128,1536,0.004368888835112254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,128,64,0.010020444790522257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,128,32,0.0025555555605226094
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,64,65536,0.010055999788973067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,128,1024,0.01036533299419615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,128,32,0.010009778042634329
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,64,16384,0.006598222172922558
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,64,12288,0.006262222097979651
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,64,65536,0.04790666699409485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,64,10240,0.005945777727497949
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,64,16384,0.020655999581019085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,64,8192,0.0069226668112807805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,64,12288,0.017869333426157635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,768,0.005945777727497949
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,64,7168,0.005929777605666055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,64,10240,0.01721333298418257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,64,6144,0.005919999960396025
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,64,5120,0.005592000153329637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,64,8192,0.016159999701711867
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,64,4096,0.005884444548024072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,64,6144,0.01482399966981676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,64,3584,0.006228444476922353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,64,5120,0.013429332938459186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,128,2560,0.01277599980433782
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,64,4096,0.013472889032628802
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,64,2560,0.006218666831652324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,64,3584,0.013421333498424955
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,64,2048,0.006239111224810283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,128,0.005713777823580637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,64,3072,0.013085333009560903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,128,64,0.0028088889602157804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,64,2560,0.012849777936935425
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,64,1024,0.0038408889538711975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,64,768,0.0034764446318149567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,64,2048,0.01239822225438224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,64,1536,0.01221244451072481
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,64,512,0.0032044444233179092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,64,256,0.0028239999794297745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,64,768,0.011689778003427716
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,64,128,0.002512000087234709
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,64,512,0.011381333072980246
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,64,64,0.0028168888141711554
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,64,256,0.011720889144473605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,64,32,0.00254488881263468
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,64,128,0.010728889041476779
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,32,65536,0.009333333207501305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,64,32,0.010001777774757808
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,32,16384,0.006234666539563074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,32,65536,0.04822399881150988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,32,12288,0.005939555664857228
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,64,7168,0.015823110938072205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,32,16384,0.020911999874644812
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,32,10240,0.005913777897755305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,32,8192,0.0052382221652401825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,32,12288,0.0185244447655148
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,64,3072,0.005575110928879843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,32,7168,0.005603555589914322
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,32,10240,0.01717511150572035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,32,6144,0.00658044425977601
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,32,8192,0.016235555211702984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,32,5120,0.0055440002017551
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,64,1536,0.004285333471165763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,32,7168,0.014466666513019137
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,32,4096,0.005598222215970357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,32,6144,0.014831999937693277
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,32,3584,0.005904888941182031
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,32,3072,0.005932444499598608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,32,5120,0.014121777481502958
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,64,1024,0.011541333463456897
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,2,128,6144,0.009701333112186855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,32,2560,0.005929777605666055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,32,4096,0.012792000340090858
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,32,2048,0.00591466658645206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,32,1536,0.004269333349333869
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,32,3584,0.013411555853154926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,32,1024,0.003681777666012446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,32,3072,0.013153778182135688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,32,768,0.0031831111345026228
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,32,2560,0.01275022245115704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,32,512,0.003173333282272021
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,32,2048,0.012423999607563019
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,32,256,0.002557333351837264
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,32,1536,0.012280888855457306
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,32,1024,0.011717333561844297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,32,128,0.002513777878549364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,32,64,0.0025439999169773525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,32,768,0.011919111013412476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,2,32,32,0.00254133323000537
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,32,512,0.011072888970375061
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,64,64,0.010033778018421596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,32,256,0.01167555567291048
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,32,64,0.009047110875447592
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,32,32,0.010029333333174387
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,65536,16384,0.334369765387641
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,65536,16384,0.2589164574940999
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,65536,12288,0.19784177674187553
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,16384,0.1597928868399726
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,65536,10240,0.21371911631690133
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,12288,0.1264293326271905
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,65536,10240,0.13449511263105604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,65536,8192,0.1750462187661065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,65536,7168,0.14548356003231472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,10240,0.11027999718983968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,8192,0.09321955839792888
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,65536,8192,0.12873510519663492
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,65536,6144,0.12563110722435847
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,65536,7168,0.12776977486080593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,65536,6144,0.09371288617451985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,7168,0.0859866672092014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,2,32,128,0.01110400011142095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,65536,5120,0.10680888758765326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,6144,0.07764710982640584
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,65536,5120,0.08260889185799493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,65536,4096,0.09051199754079182
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,65536,3584,0.07303644551171197
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,5120,0.06326666805479261
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,65536,4096,0.06926044490602282
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,65536,12288,0.2555155489179823
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,65536,3072,0.06346666812896729
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,65536,3584,0.06505511204401652
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,4096,0.05071466830041674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,65536,2560,0.05336088935534159
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,3584,0.0454364452097151
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,65536,2048,0.043967998690075345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,65536,3072,0.05918577644560072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,3072,0.04079200161827935
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,65536,2560,0.05124800072775947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,2560,0.033992889854643084
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,2048,0.028660443094041612
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,65536,1536,0.03580710954136319
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,65536,2048,0.04316177633073595
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,1536,0.02398577829202016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,65536,768,0.019151111443837483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,65536,1024,0.026518222358491685
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,65536,512,0.013637333280510373
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,768,0.014891554911931356
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,65536,768,0.023048889305856492
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,65536,256,0.008592888712882996
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,65536,512,0.01966399947802226
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,65536,128,0.006220444622966979
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,256,0.010383110907342698
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,128,0.01035911093155543
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,65536,64,0.005609777652555042
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,65536,64,0.01312711089849472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,65536,32,0.0057555557125144536
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,65536,32,0.01312711089849472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,65536,1536,0.03389511174625821
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,65536,1024,0.02381244467364417
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,1024,0.017844445175594754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,16384,65536,0.31722044944763184
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,16384,65536,0.21041599909464517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,16384,16384,0.08408800098631118
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,65536,256,0.015797333584891427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,16384,0.05338044299019707
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,16384,16384,0.07853333155314128
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,65536,0.16684088442060682
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,65536,128,0.013163555827405719
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,16384,12288,0.06074221928914388
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,16384,10240,0.05744533406363594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,12288,0.04063466522428725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,16384,8192,0.04355733262168037
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,16384,10240,0.05173866616355049
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,16384,7168,0.038252443075180054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,10240,0.03699644406636556
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,16384,8192,0.04519111249181959
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,8192,0.03103289008140564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,16384,6144,0.032893333170149065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,16384,7168,0.04525599877039591
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,16384,5120,0.02808533443344964
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,7168,0.030396444929970637
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,16384,4096,0.024614221519894067
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,16384,6144,0.03936000002755059
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,6144,0.02591111097070906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,16384,5120,0.03243555625279745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,16384,3584,0.02092622220516205
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,5120,0.02345866627163357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,16384,3072,0.01853155593077342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,16384,4096,0.028129776318868
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,4096,0.02019466625319587
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,16384,3584,0.027338667048348322
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,3584,0.01776088939772712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,16384,2560,0.016472889317406546
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,16384,3072,0.024824000067181055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,16384,2560,0.022970666488011677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,3072,0.016756445169448853
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,65536,512,0.012578667038016848
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,16384,1536,0.010661332971519895
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,2560,0.01440711153878106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,16384,2048,0.020199111766285367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,2048,0.012990222209029727
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,16384,1536,0.01856355534659492
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,1536,0.011002666420406766
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,16384,768,0.007222221957312689
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,1024,0.008960888617568547
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,16384,512,0.004917333523432414
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,16384,768,0.014781332678265043
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,768,0.00795111142926746
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,16384,256,0.003516444315512975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,16384,512,0.013748444616794586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,512,0.007310222420427535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,16384,128,0.0028106667515304354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,16384,256,0.011720000041855706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,256,0.006627555522653792
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,16384,64,0.0028702221396896574
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,16384,128,0.010682666467295753
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,16384,128,0.006617777877383762
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,16384,32,0.0028044444819291434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,16384,64,0.01068800025516086
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,16384,32,0.010593777729405297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,12288,65536,0.24111199378967285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,16384,2048,0.013784888717863293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,12288,65536,0.18699467182159424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,16384,12288,0.06522133615281847
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,16384,1024,0.008612444831265343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,12288,12288,0.051683555046717324
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,65536,0.14297866821289062
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,16384,0.04349066813786825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,12288,16384,0.06781688663694593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,16384,1024,0.015222221612930298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,12288,10240,0.04346222347683377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,12288,0.03622488843070136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,12288,10240,0.04586666822433472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,10240,0.03221688999070062
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,12288,7168,0.03509955604871114
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,8192,0.027803556786643133
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,12288,7168,0.04035022192531162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,12288,6144,0.031195554468366835
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,12288,6144,0.03558933403756883
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,6144,0.023694222172101338
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,12288,5120,0.026421333352724712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,12288,5120,0.028375112348132666
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,5120,0.020426667398876615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,12288,4096,0.023369777533743117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,12288,16384,0.06541955471038818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,4096,0.01763288842307197
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,12288,12288,0.054992887708875865
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,12288,3584,0.0248364441924625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,12288,8192,0.03863644599914551
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,12288,3072,0.018221333622932434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,3584,0.015596444408098856
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,12288,8192,0.03968266646067301
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,12288,2560,0.016000888413853116
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,3072,0.014715555641386243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,12288,2048,0.013811555173661975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,2560,0.012647999657524956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,7168,0.025068443682458665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,12288,1536,0.00889955543809467
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,12288,2048,0.01955200069480472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,2048,0.011077333655622272
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,12288,1024,0.00699733363257514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,1536,0.009826666778988307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,12288,1536,0.017481777403089736
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,12288,768,0.005873777800136142
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,12288,1024,0.014694222145610385
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,1024,0.008336000144481659
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,12288,512,0.0040835556056764396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,12288,768,0.014614222778214348
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,768,0.007673777639865875
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,12288,256,0.0031795555518733132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,12288,4096,0.02604622145493825
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,12288,512,0.01312711089849472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,512,0.0069475554757648045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,12288,128,0.002842666581273079
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,12288,256,0.011039111349317761
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,256,0.006633777585294511
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,12288,64,0.0028373334142896864
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,12288,128,0.006586666736337874
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,12288,32,0.002793777734041214
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,12288,64,0.010041777458455827
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,12288,32,0.010398222340477837
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,12288,3072,0.022963555322753057
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,10240,65536,0.2036231093936496
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,12288,2560,0.02094488839308421
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,10240,65536,0.1698737806744046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,10240,16384,0.05443911088837517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,10240,12288,0.04371466570430332
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,16384,0.043015112479527794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,65536,0.13828889528910318
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,10240,10240,0.03801066676775614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,10240,12288,0.0443173348903656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,12288,0.033406221204333834
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,10240,10240,0.03965333435270522
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,10240,8192,0.03174488743146261
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,10240,0.03016800019476149
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,10240,7168,0.029430223835839167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,10240,8192,0.034394668208228216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,8192,0.02497244377930959
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,12288,3584,0.020029332902696397
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,10240,6144,0.02586488922437032
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,10240,7168,0.03014311194419861
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,7168,0.02360711163944668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,10240,5120,0.022250667214393616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,10240,6144,0.027456000447273254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,6144,0.020847999387317236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,12288,128,0.010440000229411654
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,10240,3584,0.018797333041826885
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,10240,5120,0.02441155579355028
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,4096,0.01588888963063558
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,10240,3072,0.01647644407219357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,3584,0.015032000011867948
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,10240,16384,0.056551112069023975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,3072,0.013785777820481194
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,10240,2560,0.014923555983437432
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,10240,2560,0.018794667389657762
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,2560,0.012083555261294046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,10240,2048,0.010996444357766045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,10240,2048,0.017651556266678702
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,2048,0.01036622209681405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,10240,1536,0.008943999807039896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,10240,1536,0.01610488858487871
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,1536,0.009366222553782994
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,10240,1024,0.006550222221348021
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,10240,1024,0.013783999615245394
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,1024,0.007985778152942657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,10240,768,0.00453333349691497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,10240,4096,0.01900622248649597
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,10240,512,0.0038648889296584656
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,10240,768,0.013775111072593264
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,768,0.007289778027269576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,10240,4096,0.022449778185950384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,10240,256,0.0031555555760860443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,10240,3584,0.022642667094866436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,10240,512,0.01314755529165268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,512,0.006888000087605582
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,10240,128,0.002874666617976295
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,10240,64,0.0028328889360030494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,256,0.0063999998900625445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,10240,256,0.011494222614500256
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,10240,3072,0.020230222079488967
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,10240,32,0.002510222295920054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,128,0.006567111031876669
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,10240,128,0.010070222119490305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,10240,64,0.010384000009960597
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,10240,32,0.010399111443095738
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,8192,16384,0.04620088802443611
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,8192,65536,0.16784800423516166
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,8192,12288,0.0344471103615231
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,8192,65536,0.11758489078945583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,8192,16384,0.043828444348441235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,16384,0.024846222665574815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,8192,10240,0.03050133254792955
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,65536,0.0782453351550632
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,8192,8192,0.025215999947653875
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,8192,12288,0.0358515547381507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,8192,10240,0.031932443380355835
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,8192,7168,0.021746666895018682
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,8192,8192,0.02681066592534383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,8192,0.017312000195185345
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,8192,6144,0.018776888648668926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,8192,7168,0.027267555395762127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,7168,0.015459555718633862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,8192,5120,0.017130666308932833
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,6144,0.014923555983437432
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,8192,4096,0.013952000273598565
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,10240,5120,0.019112888309690688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,8192,5120,0.020928889513015747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,5120,0.01349244432316886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,8192,3584,0.013150222599506378
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,4096,0.011359111302428775
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,8192,3584,0.019889778561062284
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,8192,3072,0.011855111353927188
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,3584,0.010515555739402771
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,3072,0.009703111317422655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,8192,3072,0.018226666582955253
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,8192,2560,0.00997866690158844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,2560,0.008965333302815756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,8192,2048,0.008790222307046255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,12288,0.020670221911536325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,8192,1536,0.007354666789372762
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,8192,2048,0.015783111254374187
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,2048,0.008569777839713627
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,8192,1024,0.0058204444746176405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,8192,1536,0.01480888823668162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,1536,0.007272889216740926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,8192,768,0.0041893331540955436
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,8192,1024,0.013088000317414602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,1024,0.006657777975002925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,8192,512,0.0038604442444112566
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,8192,768,0.012774222426944308
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,768,0.006620444357395172
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,8192,256,0.0032035555276605818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,8192,4096,0.018591110905011494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,8192,512,0.012406222522258759
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,512,0.006223111102978389
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,8192,128,0.0030213333666324615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,8192,256,0.011426666544543372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,256,0.005943111247486538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,8192,64,0.0028071111689011254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,8192,128,0.01016622202263938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,8192,32,0.002855111120475663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,8192,64,0.008998221821255153
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,8192,32,0.010046222143703038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,7168,65536,0.14524355199601915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,7168,16384,0.04216622312863668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,8192,2560,0.017542221479945712
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,7168,65536,0.10384711292054917
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,7168,12288,0.033933334880405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,16384,0.02272266646226247
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,65536,0.07403200202518039
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,7168,16384,0.04260711206330193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,7168,10240,0.028567999601364136
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,7168,12288,0.033857779370413885
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,7168,8192,0.0251857770813836
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,7168,10240,0.03050222330623203
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,10240,0.01625866691271464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,7168,7168,0.020224889119466145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,7168,8192,0.02602311140961117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,8192,0.014476444986131458
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,7168,6144,0.016788444585270353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,7168,7168,0.025047111842367385
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,7168,0.013437333206335703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,7168,5120,0.0148062225845125
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,6144,0.012716444830099741
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,128,0.005923555543025334
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,7168,4096,0.013429332938459186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,7168,5120,0.01991911066903008
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,5120,0.011446221835083432
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,7168,3584,0.011972444752852121
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,7168,4096,0.019241778386963736
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,4096,0.009730666875839233
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,7168,3072,0.010321777727868822
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,7168,3584,0.018631110588709515
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,8192,10240,0.018416888184017606
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,7168,2560,0.00926755534278022
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,7168,3072,0.016560888952679105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,3072,0.008674666285514832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,7168,2048,0.008277333445019191
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,12288,0.018479110466109384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,7168,2560,0.01648977730009291
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,8192,6144,0.02347822156217363
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,7168,2048,0.015106666419241162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,2048,0.007592000067234039
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,7168,1024,0.005072000126043956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,7168,1536,0.014094221923086377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,7168,1024,0.012506666282812754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,7168,768,0.0041999999019834726
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,7168,6144,0.02330044408639272
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,7168,768,0.01311466677321328
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,7168,512,0.0038364442686239877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,7168,512,0.012497777740160624
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,3584,0.009690666364298927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,7168,256,0.003181333343187968
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,512,0.005952888892756567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,7168,256,0.010427555276287926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,256,0.005888000130653381
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,7168,128,0.0028355556229750314
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,2560,0.00798044436507755
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,7168,64,0.0028453332682450614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,7168,1536,0.006923555499977536
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,7168,32,0.0028479999552170434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,128,0.005671111245950063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,7168,64,0.009309333231714036
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,7168,32,0.010060444474220276
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,6144,65536,0.12594755490620932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,6144,16384,0.03879822293917338
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,6144,65536,0.09856888982984756
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,1024,0.006583111153708563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,65536,0.07202488846249051
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,6144,12288,0.030449777841567993
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,6144,16384,0.04255200094646878
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,16384,0.021910222040282354
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,6144,10240,0.025581333372328017
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,768,0.00628355559375551
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,6144,12288,0.033230221933788724
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,12288,0.01815555493036906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,6144,8192,0.021976888179779053
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,6144,10240,0.02779910961786906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,10240,0.016032889485359192
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,6144,7168,0.020050666398472257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,8192,0.014261333478821648
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,6144,6144,0.01792266633775499
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,6144,7168,0.025715554753939312
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,6144,5120,0.015862221519152325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,6144,6144,0.022105778257052105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,5120,0.01167022188504537
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,6144,5120,0.01981599960062239
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,6144,4096,0.012313777373896705
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,7168,128,0.01036000003417333
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,6144,3584,0.010808889236715106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,6144,4096,0.018576888574494254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,6144,3584,0.018896889355447557
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,3584,0.009015111459626092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,6144,3072,0.009686222506894005
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,3072,0.008444444172912175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,6144,3072,0.017064889272054035
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,6144,2560,0.008653333617581261
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,6144,2560,0.016873778568373788
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,2560,0.00794488853878445
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,6144,2048,0.007289778027269576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,6144,8192,0.025331555141343012
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,6144,1536,0.006211555666393704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,6144,2048,0.014481777946154276
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,2048,0.007675555845101674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,6144,1024,0.004388444539573458
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,6144,0.012150222228633033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,1536,0.006815111057625876
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,6144,1536,0.013798221945762634
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,6144,768,0.0038942222793896994
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,6144,1024,0.01275111072593265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,6144,512,0.00328533330725299
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,1024,0.006562666760550604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,6144,768,0.0127697777416971
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,6144,256,0.0031555555760860443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,6144,512,0.012127111355463663
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,7168,1536,0.006617777877383762
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,6144,128,0.0028044444819291434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,6144,256,0.011006222003036074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,256,0.005931555396980709
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,6144,64,0.0028497777465316984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,128,0.0058204444746176405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,6144,128,0.010043555663691627
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,6144,32,0.002514666567246119
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,6144,64,0.009340444372759925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,6144,32,0.010037333601050908
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,5120,65536,0.10679288705190022
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,5120,16384,0.03261422117551168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,5120,65536,0.09023555782106187
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,65536,0.06919911172654894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,7168,0.013424889081054263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,5120,12288,0.02698844505680932
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,16384,0.02125066684352027
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,5120,10240,0.02308622168170081
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,5120,12288,0.029654221402274236
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,12288,0.017271111408869427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,5120,8192,0.01999200052685208
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,5120,10240,0.025832888152864244
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,10240,0.015463999576038785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,768,0.00630044440428416
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,5120,8192,0.024311999479929607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,8192,0.013360000318951078
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,512,0.005947555518812603
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,5120,7168,0.025077333052953083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,7168,0.012650666965378655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,5120,5120,0.014488000008794995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,5120,6144,0.02110133402877384
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,6144,0.011388444238238864
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,5120,4096,0.010947555303573608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,5120,5120,0.01907466683122847
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,5120,0.010245333115259806
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,5120,3584,0.00963644435008367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,5120,4096,0.018202667435010273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,4096,0.008634666601816813
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,5120,3072,0.0084915558497111
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,5120,3584,0.017882666654056974
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,3584,0.008536888493431939
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,5120,2560,0.00795999997191959
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,5120,3072,0.016137777103318107
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,3072,0.007963555554548899
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,5120,2048,0.0069013333155049225
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,5120,2560,0.015491555134455362
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,2560,0.007353777686754863
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,5120,1536,0.005437333136796951
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,2048,0.007101332975758447
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,5120,2048,0.014129777749379476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,5120,16384,0.03906044363975525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,5120,1024,0.0041866666740841335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,5120,1536,0.01275111072593265
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,5120,768,0.003549333247873518
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,1024,0.006288888967699475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,5120,1024,0.012423999607563019
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,5120,512,0.003177777760558658
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,5120,768,0.011773332953453064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,768,0.006240889016124938
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,5120,7168,0.018238221605618794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,5120,256,0.00286666676402092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,5120,512,0.012030222349696688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,512,0.005921777751710679
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,5120,128,0.0030382221771611106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,6144,4096,0.010031110710567897
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,256,0.00591111100382275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,5120,64,0.002789333255754577
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,5120,32,0.002524444419476721
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,128,0.005856888989607493
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,5120,128,0.009343999955389235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,5120,64,0.010033778018421596
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,4096,65536,0.08715999788708156
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,5120,32,0.009973333113723332
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,4096,16384,0.02787999974356757
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,4096,65536,0.08645244439442952
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,4096,12288,0.02100177771515316
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,4096,16384,0.03220444586541917
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,65536,0.06698755423227946
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,16384,0.020447999238967896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,4096,10240,0.01867911054028405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,12288,0.01679999960793389
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,4096,12288,0.03011288907792833
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,4096,10240,0.024467556013001338
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,4096,8192,0.016787555482652452
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,10240,0.014925332532988654
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,4096,7168,0.014799111419253878
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,5120,1536,0.006924444602595435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,4096,8192,0.023600889576805964
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,8192,0.013033777475357056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,4096,7168,0.02148799929353926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,7168,0.011957333319716983
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,4096,5120,0.01219555570019616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,6144,0.010727110836240979
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,4096,4096,0.010696888797812991
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,4096,5120,0.01833599971400367
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,5120,0.009672889278994666
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,5120,6144,0.016186666157510545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,4096,4096,0.016799110505315993
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,4096,0.008625778059164682
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,4096,3072,0.010474666953086853
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,4096,3584,0.016472889317406546
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,3584,0.008271999657154083
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,4096,2560,0.010019555687904358
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,4096,3072,0.015125332607163323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,3072,0.007962666451931
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,4096,2048,0.007651555869314406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,4096,2560,0.014076444837782117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,2560,0.0075777777367168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,4096,1536,0.006911110960774952
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,4096,2048,0.013750221994188098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,2048,0.007185777856243982
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,4096,1024,0.0052328887912962176
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,4096,1536,0.013068444199032254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,4096,1024,0.012092444631788464
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,4096,6144,0.013396444420019785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,4096,768,0.004604444321658877
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,768,0.006227555374304454
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,4096,768,0.012449777788586087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,4096,512,0.0038702223036024305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,4096,3584,0.012695999609099494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,4096,512,0.012081777883900536
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,512,0.005929777605666055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,5120,256,0.010716444088353051
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,4096,128,0.0028773333049482773
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,256,0.005863111052248213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,4096,256,0.011124444504578909
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,4096,64,0.002516444358560774
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,4096,128,0.009687999884287516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,4096,32,0.0028497777465316984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,4096,64,0.010047111246320937
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,4096,32,0.009400889277458191
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3584,65536,0.07929599947399564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,1536,0.0069146665434042616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3584,65536,0.08123555448320177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3584,16384,0.025306666890780132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,65536,0.06606933143403795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,1024,0.006251555350091722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3584,12288,0.019868445065286424
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3584,16384,0.033570667107899986
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,16384,0.02018933329317305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3584,10240,0.017567111386193168
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,4096,6144,0.021772444248199463
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3584,12288,0.02756888833310869
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,12288,0.016512889001104567
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3584,8192,0.01513777838812934
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3584,10240,0.023082666926913794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3584,7168,0.013795555465751223
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,10240,0.014967999524540372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,4096,256,0.003149333306484752
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3584,8192,0.022610666023360357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3584,6144,0.01308444473478529
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,7168,0.011714666253990598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3584,5120,0.011694221860832639
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3584,6144,0.01886133352915446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,6144,0.010778666370444827
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3584,4096,0.010241777532630498
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,4096,128,0.005703111075692707
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,5120,0.009546666509575313
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3584,3584,0.012030222349696688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3584,4096,0.016581333345837064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3584,3072,0.010015111002657149
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,3584,0.008143111235565608
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,3072,0.007655111451943715
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3584,2560,0.009694221946928237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3584,2560,0.014819555812411837
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,2560,0.007311111523045435
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3584,2048,0.008279110822412703
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,8192,0.012655110822783576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3584,7168,0.020232000284724765
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,2048,0.007263110743628607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3584,1536,0.005962666538026597
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,1536,0.0069253332912921906
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3584,1536,0.01312355531586541
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3584,5120,0.017515555024147034
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3584,1024,0.005220444252093633
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3584,768,0.004204444587230682
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,4096,0.008610666626029545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3584,1024,0.012063110868136087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3584,512,0.00388355553150177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3584,3584,0.015133332875039844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,768,0.006027555714050929
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3584,256,0.0028657778683635923
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3584,3072,0.01482399966981676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3584,256,0.011048888994587792
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,256,0.0059279998143513995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3584,128,0.0027991111079851785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,128,0.0058782220714622075
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3584,128,0.009691555466916826
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3584,64,0.0027840000887711844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3584,32,0.002571555475393931
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3584,64,0.009331555830107795
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3584,2048,0.013779555757840475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3072,65536,0.06773066520690918
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3072,65536,0.07221777571572198
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3072,16384,0.021585777401924133
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,65536,0.06518755356470744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,1024,0.006260444306664997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3072,12288,0.017889777819315594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3584,768,0.012387555506494312
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3072,16384,0.029310223129060533
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3072,10240,0.015433778365453085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3584,512,0.012379555238617791
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3072,12288,0.02665155629316966
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3072,8192,0.013567999833159976
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3072,10240,0.022623111804326374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,10240,0.014459555347760519
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3072,7168,0.012805333567990197
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3072,8192,0.021287999219364587
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,8192,0.01238933371173011
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3072,6144,0.011791999969217511
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3072,7168,0.020444444484180875
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,7168,0.011095110740926532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3072,5120,0.010709332923094431
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3072,6144,0.018560000591807894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,6144,0.010899555351999072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3584,32,0.00999111102686988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3072,4096,0.009870222045315636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3072,5120,0.016185777054892648
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3072,3584,0.009319110876984065
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,5120,0.009693333672152625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,4096,0.008609777523411645
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3072,4096,0.0151591118839052
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3072,3072,0.00923733330435223
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,3584,0.008310221963458592
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3072,3584,0.01479555500878228
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3072,2560,0.008346666892369589
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3072,3072,0.01407199932469262
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,3072,0.00794577764140235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3072,2048,0.007245333658324347
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3072,2560,0.013097777962684631
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,2560,0.007663999994595845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3072,1536,0.007176888485749562
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3072,2048,0.012796444197495779
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3072,1024,0.005254222287072076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3072,1536,0.01276177747382058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,1536,0.006613333192136552
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3072,768,0.004943999979231092
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3072,1024,0.011701333026091257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3072,768,0.011393778026103973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3072,512,0.0038817777401871155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,768,0.005958222266700532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3072,256,0.0034871109657817413
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3072,512,0.01105955574247572
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3072,256,0.011056889262464313
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3072,128,0.0031635556370019913
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3072,128,0.010054222411579555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,128,0.005616000129116907
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3072,64,0.002524444419476721
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,16384,0.019973332683245342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,3072,32,0.0027555556346972785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,2048,0.007264888948864407
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,3584,512,0.005940444353553984
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2560,65536,0.05886399745941162
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3072,32,0.00999111102686988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,1024,0.006288888967699475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2560,16384,0.01977066695690155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2560,65536,0.06898488601048787
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,65536,0.061905774805280894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2560,12288,0.015480000111791821
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2560,16384,0.028252445989184912
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,512,0.005933333188295364
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2560,10240,0.014404444230927361
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,16384,0.019504888190163504
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2560,12288,0.0240320000383589
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,12288,0.015896888242827523
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2560,8192,0.012160888976520963
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,10240,0.013744889034165276
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2560,7168,0.011462222370836469
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2560,8192,0.020290666156344943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,8192,0.011391999820868174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2560,6144,0.010636444720957013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2560,7168,0.01848266687658098
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,7168,0.010661332971519895
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2560,5120,0.0099653336736891
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2560,6144,0.016672000288963318
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,6144,0.00999022192425198
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2560,4096,0.008943110704421997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,3072,64,0.010054222411579555
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2560,5120,0.015306666493415833
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,5120,0.009324444664849175
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,12288,0.01633777717749278
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2560,3584,0.008270222279760573
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2560,3072,0.008658666577604082
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,4096,0.008637333081828224
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2560,4096,0.01384800010257297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2560,3584,0.013502221968438892
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,3584,0.00832799987660514
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,3072,0.007928888830873702
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2560,2048,0.006616888774765863
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,2560,0.007640000018808577
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2560,2048,0.012484444512261285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,2048,0.007300444775157505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2560,1536,0.006605333338181178
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,3072,256,0.005578666511509154
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2560,1024,0.004895110925038655
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2560,10240,0.0214684440029992
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,1536,0.0069102222720781965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2560,768,0.004572444491916233
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2560,1024,0.01071911139620675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2560,768,0.01238222254647149
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2560,512,0.003958222352796131
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,768,0.0059164443777667145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2560,512,0.012100444071822695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,512,0.0058053334554036455
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2560,256,0.0034808889031410217
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2560,256,0.010046222143703038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2560,2560,0.007631111476156447
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2560,128,0.003141333245568805
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2560,3072,0.013435555828942193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2560,64,0.0028239999794297745
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,128,0.005608888963858287
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2560,128,0.010064889159467485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2560,32,0.00254488881263468
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2560,64,0.01001422190003925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2560,32,0.009986666341622671
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2560,1536,0.0127688886390792
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2048,65536,0.07301244470808241
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,65536,0.059747556845347084
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,1024,0.006279110908508301
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,16384,0.01918577816751268
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2048,12288,0.02274488906065623
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,12288,0.01518133282661438
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2048,10240,0.012079111403889127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,2560,256,0.005591111050711737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2048,10240,0.02046133412255181
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,10240,0.012899555265903473
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2048,8192,0.010945777926180096
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2560,2560,0.012871111432711283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2048,7168,0.010616888602574667
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2048,8192,0.018627555833922494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,8192,0.01109155515829722
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2048,6144,0.009717333647939894
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2048,7168,0.017316444052590262
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,7168,0.010634666515721215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2048,5120,0.008737777670224508
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2048,6144,0.015490666031837463
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2048,16384,0.01646933290693495
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2048,4096,0.010325333310498131
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2048,5120,0.01480977733929952
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,5120,0.009299555586444007
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2048,16384,0.0281057788266076
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2048,3584,0.009679999616410997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2048,4096,0.014249778456158109
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,4096,0.008296888735559251
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2048,3584,0.013776000175211163
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2048,3072,0.013419555293189155
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,3072,0.007792000141408708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2048,2560,0.013427555561065674
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,2560,0.007345777418878343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2048,2048,0.007975111405054728
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2048,2048,0.012518222133318583
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2048,65536,0.05000177688068814
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,2048,0.007271111011505127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2048,1536,0.0069448889957533936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2048,1536,0.01239999963177575
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,1536,0.006614222294754452
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,6144,0.009975999593734741
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,1024,0.006253333141406377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2048,12288,0.01330844478474723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2048,768,0.004217777815130022
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2048,3072,0.007241778075695038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2048,512,0.003906666818592283
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,3584,0.00794577764140235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2048,2560,0.006562666760550604
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,512,0.005722666780153911
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2048,256,0.0031866667171319327
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2048,256,0.010994666980372535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,256,0.005895111295912001
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2048,128,0.00281866660548581
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,128,0.005564444594913059
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2048,128,0.009352000223265754
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2048,64,0.002798222212327851
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2048,32,0.0028595555987623003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2048,64,0.009343110852771336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2048,32,0.00999111102686988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1536,65536,0.03546488947338528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,2048,1024,0.004869333157936732
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1536,16384,0.014008000493049622
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1536,65536,0.05879200167126126
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2048,1024,0.010716444088353051
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,65536,0.0590933362642924
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1536,16384,0.023660444551044043
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,16384,0.018353778454992507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,2048,768,0.006048000107208888
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1536,10240,0.010498666928874122
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,12288,0.01386755539311303
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1536,12288,0.022301332818137273
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1536,8192,0.010267555713653564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2048,512,0.01238222254647149
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1536,10240,0.018903111418088276
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,10240,0.012715555727481842
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1536,7168,0.009677333136399588
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,8192,0.011023999916182624
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1536,8192,0.01683466633160909
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,7168,0.01034755590889189
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1536,7168,0.015545777148670621
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1536,5120,0.007659555309348636
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1536,6144,0.015127110812399121
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1536,4096,0.007275555696752336
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1536,5120,0.01443733274936676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,5120,0.009299555586444007
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1536,3584,0.006855999843941794
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1536,4096,0.013576000101036496
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,4096,0.008328888979223039
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1536,3072,0.006266666783226862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1536,3584,0.01346666696998808
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,3584,0.0079644446571668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1536,2560,0.0057297779454125305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,3072,0.007654222349325816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1536,3072,0.013212444053755866
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1536,2048,0.006583111153708563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1536,2560,0.012721777790122561
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,2560,0.007313778003056844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1536,12288,0.011446221835083432
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1536,1536,0.005598222215970357
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1536,2048,0.012352888782819113
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,2048,0.007075555622577667
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1536,1024,0.004555555681387584
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,2048,768,0.012042666474978128
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1536,768,0.003899555653333664
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1536,1536,0.011383111278216044
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,1024,0.006258666515350342
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1536,512,0.003478222423129611
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,768,0.005968889014588461
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1536,768,0.012044444680213928
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1536,256,0.0031537777847713898
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1536,512,0.011733333269755045
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,512,0.005743999862008625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1536,128,0.0028479999552170434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1536,256,0.011005333728260465
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,256,0.005617777920431561
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1536,64,0.002789333255754577
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1536,128,0.009329777624871995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,128,0.005597333527273602
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1536,32,0.0028008888992998335
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1536,64,0.01017777787314521
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1024,65536,0.026745777991082933
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1536,32,0.01000088867213991
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1024,16384,0.015009777413474189
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1024,65536,0.05681066380606758
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,65536,0.058268447717030845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1024,12288,0.012800888882742988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1024,16384,0.022304000125990972
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,16384,0.01683999929163191
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1024,10240,0.011062222222487131
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1024,12288,0.01919644408755832
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,12288,0.01388711151149538
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1024,8192,0.010437332921557956
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1024,10240,0.017497777938842773
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1536,1024,0.011412444214026133
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,10240,0.012416889270146688
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1024,7168,0.009357333183288574
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,8192,0.011036444041464063
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1024,8192,0.01641688909795549
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1024,6144,0.008622222476535374
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1024,7168,0.015180443723996481
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,7168,0.010378667049937779
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,6144,0.010016889207892949
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1024,5120,0.007978666987684038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1024,6144,0.015438222222858004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,6144,0.010031999813185798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1024,4096,0.007303111255168915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,5120,0.009332444104883406
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1024,3584,0.006897777732875612
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1024,4096,0.013436444103717804
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1024,3072,0.006241777704821692
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1024,3584,0.013448889056841532
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,3584,0.00795199970404307
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1024,2560,0.0059004442559348206
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1024,3072,0.012488000094890594
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,3072,0.007639110916190677
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1024,2048,0.0075786668393347
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1024,2560,0.013405332962671915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1024,2048,0.012353777885437012
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,2048,0.006970666348934174
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1024,1536,0.006263111200597551
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,1536,1536,0.006629333313968446
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1024,1024,0.00490399988161193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,1536,0.006621333460013072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1536,6144,0.008949333594905006
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1024,1024,0.011413333316644033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,1024,0.006267555471923616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1024,768,0.0120319997270902
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,768,0.006233777850866318
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1024,512,0.0035253332720862497
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1024,5120,0.014527999692493014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1024,256,0.0031795555518733132
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1024,512,0.011737777955002256
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,512,0.005902222047249476
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,4096,0.00832177781396442
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1024,128,0.0028844444702068963
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,256,0.005559111220969095
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1024,256,0.010382222632567087
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,128,0.005581333405441708
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1024,128,0.00999022192425198
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1024,32,0.002814222127199173
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,768,65536,0.02241244415442149
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1024,64,0.009354666703277165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1024,32,0.009736888938479954
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,1024,2560,0.007313778003056844
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,768,16384,0.010242666635248397
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,768,65536,0.05225777626037598
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,65536,0.057968888017866343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,768,12288,0.010833777487277985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,768,16384,0.02089599933889177
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,16384,0.01680799987581041
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,768,10240,0.009680888719028896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,12288,0.013849777479966482
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,1024,1536,0.012503999802801343
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,768,8192,0.008982222113344405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,768,10240,0.01755199995305803
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,768,7168,0.008943110704421997
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1024,768,0.004535111288229625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,768,8192,0.016197333733240765
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,8192,0.011010666688283285
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,768,6144,0.007978666987684038
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,768,7168,0.015844444433848064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,7168,0.010398222340477837
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,768,5120,0.006911110960774952
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,768,6144,0.015122666954994202
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,6144,0.010004444254769219
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,768,4096,0.007245333658324347
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,5120,0.009305777649084726
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,768,3584,0.0069520001610120135
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,4096,0.008287111090289222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,1024,64,0.0028435554769304064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,768,3584,0.014184888866212634
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,768,3072,0.006605333338181178
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,3072,0.007655999726719326
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,768,2560,0.006214222146405115
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,768,2560,0.012730666332774691
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,768,12288,0.01854933301607768
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,10240,0.012392000191741519
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,2560,0.007300444775157505
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,2048,0.006950221955776215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,768,1536,0.005541333307822545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,768,1536,0.012366222010718452
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,768,5120,0.014459555347760519
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,768,1024,0.0042364444169733255
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,768,4096,0.013912000589900546
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,1536,0.006609777609507243
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,768,768,0.003882666842805015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,768,1024,0.011700444751315646
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,3584,0.007996444072988298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,768,512,0.003504000190231535
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,768,768,0.0120328888297081
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,768,0.005982222242487802
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,768,3072,0.013413333230548434
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,768,256,0.0031662223239739737
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,768,512,0.011741333537631564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,512,0.0058871110280354815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,768,128,0.002790222151411904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,768,256,0.010976889067225985
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,768,128,0.010040889183680216
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,768,64,0.002514666567246119
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,768,2048,0.011398221883508893
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,768,32,0.002842666581273079
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,768,32,0.00905600024594201
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,512,65536,0.01716711123784383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,1024,0.006255110932721033
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,65536,0.051744000779257886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,512,16384,0.010989333192507425
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,512,16384,0.020815999971495736
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,16384,0.01683022247420417
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,512,12288,0.008743111458089616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,256,0.005592888842026393
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,512,10240,0.007970666719807519
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,512,12288,0.018578666779730055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,768,128,0.005579555614127054
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,768,2048,0.005244444641802046
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,10240,0.01274311128589842
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,512,8192,0.016181333197487723
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,8192,0.011016888750924004
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,512,7168,0.007631111476156447
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,512,7168,0.015807110402319167
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,7168,0.010384000009960597
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,512,6144,0.007304000357786815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,512,6144,0.01516711049609714
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,512,5120,0.0069777775141927935
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,512,65536,0.051550222767723926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,512,5120,0.014398222168286642
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,5120,0.009020444419648912
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,512,4096,0.0069653333889113525
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,512,4096,0.013090666797426013
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,4096,0.00830666638082928
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,512,3584,0.0069422221018208405
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,12288,0.013754666679435305
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,512,3072,0.0063164445261160536
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,512,3584,0.01366666704416275
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,512,10240,0.017502221796247695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,512,2560,0.006579555571079254
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,512,3072,0.013088888592190213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,3072,0.007638221813572778
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,768,64,0.010038221875826517
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,512,2048,0.005938666562239329
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,2560,0.007295110987292395
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,512,2560,0.01313866674900055
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,512,1536,0.0047608887155850725
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,2048,0.006962666908899943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,512,1024,0.003925333420435588
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,512,1536,0.012396444049146442
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,1536,0.006908444480763541
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,512,768,0.0034888887570963967
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,6144,0.009855999714798397
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,1024,0.006285333385070165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,512,1024,0.011720889144473605
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,512,512,0.003305777700410949
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,512,768,0.0120328888297081
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,768,0.006238222122192383
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,512,512,0.011931555966536203
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,512,0.005674666828579373
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,512,128,0.0028293333533737394
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,512,256,0.011017777853541903
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,256,0.005575110928879843
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,512,64,0.0025173332542181015
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,512,128,0.01034222212102678
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,128,0.005602666487296422
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,512,32,0.002806222273243798
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,256,65536,0.015807999504937064
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,512,64,0.009343999955389235
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,512,8192,0.007972444097201029
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,512,32,0.0099973330895106
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,256,16384,0.008953777452309927
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,256,65536,0.049250665638181895
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,65536,0.05047466688685947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,256,12288,0.007993777592976889
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,256,16384,0.02090666691462199
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,16384,0.01679644485314687
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,256,10240,0.007600888609886169
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,256,12288,0.018536888890796237
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,512,2048,0.01241422196229299
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,256,8192,0.00665244460105896
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,256,10240,0.017367111312018502
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,256,7168,0.006317333214812809
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,8192,0.011054221954610614
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,256,7168,0.015676443775494892
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,256,6144,0.006228444476922353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,512,256,0.0031679999083280563
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,256,6144,0.015293333265516492
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,256,5120,0.006065777606434292
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,256,5120,0.014436443646748861
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,256,4096,0.005531555662552516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,5120,0.008995555341243744
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,512,3584,0.007968000239796108
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,256,3584,0.005910222315125995
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,4096,0.008316444853941599
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,256,4096,0.013787555197874704
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,256,3072,0.0058871110280354815
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,256,3584,0.012984000146389008
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,3584,0.008007999923494127
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,12288,0.013791110780504016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,256,2560,0.006599999964237213
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,3072,0.007613333563009898
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,10240,0.012714666624863943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,256,2048,0.005931555396980709
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,2560,0.007267555428875818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,256,8192,0.016057777735922072
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,256,1536,0.004219555606444676
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,256,2048,0.012154666913880242
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,2048,0.006978666616810693
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,7168,0.01036533299419615
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,256,1024,0.0038720000949170855
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,256,1536,0.012367111113336353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,1536,0.006631111105283101
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,256,768,0.0033200000309281876
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,256,1024,0.011741333537631564
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,256,512,0.0031902222997612427
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,768,0.006249777972698212
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,256,512,0.011722666521867117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,512,0.005686222265164058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,256,256,0.0028684443483750024
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,256,0.005574222240183089
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,256,128,0.002838222309947014
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,256,128,0.009405333134863112
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,256,3072,0.013122667041089801
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,256,64,0.0024684444069862366
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,128,0.005601777798599667
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,256,32,0.0025902222841978073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,256,2560,0.013234666652149625
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,256,64,0.009856888817416297
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,128,65536,0.014134221606784396
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,256,32,0.009347555538018545
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,128,16384,0.007333333293596904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,65536,0.049977779388427734
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,1024,0.00629688882165485
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,128,16384,0.02064622276359134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,128,12288,0.007281777759393056
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,16384,0.01683999929163191
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,128,10240,0.0069511110583941145
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,128,12288,0.018562666243977018
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,128,10240,0.017516444126764934
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,10240,0.01275288893116845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,128,8192,0.006792888873153263
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,256,256,0.010997333460383944
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,128,7168,0.006588444527652528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,8192,0.011007111105653973
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,128,8192,0.016166221764352586
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,128,6144,0.005939555664857228
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,128,7168,0.015471110741297403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,7168,0.01037688884470198
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,128,5120,0.005920889063013925
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,6144,0.010020444790522257
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,128,6144,0.015141333142916361
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,128,4096,0.0052435555391841466
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,128,5120,0.014281777871979607
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,5120,0.00924444446961085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,128,3584,0.006144000010357962
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,128,4096,0.013774221969975365
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,4096,0.008354666332403818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,128,3072,0.00592533333433999
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,128,65536,0.048575109905666776
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,128,3584,0.013435555828942193
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,3584,0.007996444072988298
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,256,6144,0.010007999837398529
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,128,3072,0.01312355531586541
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,256,768,0.01219555570019616
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,128,2048,0.005943999936183293
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,2560,0.007272889216740926
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,128,1536,0.004187555362780889
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,12288,0.01369511087735494
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,2048,0.007038222418891058
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,128,2048,0.012111999922328524
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,1536,0.006899555524190267
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,128,1536,0.012400888734393649
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,128,768,0.0034862222770849862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,1024,0.006247111078765657
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,128,512,0.003156444471743372
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,768,0.006064888917737537
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,128,768,0.012042666474978128
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,128,256,0.002817777709828483
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,128,512,0.01204711116022534
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,128,256,0.011028444601429833
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,128,128,0.0026462222966882917
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,256,0.005562666803598404
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,128,0.005552000055710475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,128,64,0.002528000002106031
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,128,64,0.00999111102686988
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,128,2560,0.006231110956933763
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,128,32,0.0027964444210131965
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,64,65536,0.011042666931947073
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,128,2560,0.01310488912794325
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,128,32,0.010023111270533668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,64,16384,0.0069448889957533936
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,128,1024,0.0038408889538711975
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,64,12288,0.006270222365856171
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,64,10240,0.005867555737495422
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,64,16384,0.020632000433074105
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,64,8192,0.006628444625271692
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,128,1024,0.011371555427710215
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,64,10240,0.017323555217848886
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,64,8192,0.015857777661747403
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,512,0.005899555567238066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,64,6144,0.0063457778758472866
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,64,7168,0.01532977819442749
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,64,5120,0.005931555396980709
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,64,6144,0.01516533394654592
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,64,4096,0.005548444473081165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,128,128,0.010056888891590966
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,64,5120,0.014152889450391134
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,64,3584,0.005960888746711943
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,64,3072,0.005951111101441913
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,64,4096,0.013805333111021252
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,64,3584,0.013445333474212222
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,64,3072,0.01312711089849472
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,64,2048,0.006228444476922353
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,nvfp4,1,128,3072,0.007654222349325816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,64,2560,0.012771555946932899
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,64,1536,0.004524444540341695
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,64,1024,0.0038275557259718576
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,64,2048,0.012400888734393649
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,64,65536,0.04823644293679131
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,64,1536,0.012324444121784635
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,64,1024,0.011361777782440186
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,64,512,0.0032275555034478507
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,64,256,0.0028364445186323593
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,64,12288,0.01755911111831665
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,64,768,0.012042666474978128
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,64,128,0.002867555452717675
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,64,7168,0.00655377780397733
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,64,512,0.011720000041855706
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,64,256,0.010716444088353051
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,64,32,0.00249422217408816
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,64,128,0.010091555615266165
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,32,65536,0.00925155563486947
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,64,64,0.010023111270533668
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,32,16384,0.006073777874310811
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,32,12288,0.005948444621430502
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,32,65536,0.04805955621931288
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,32,16384,0.020592000749376085
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,32,10240,0.006292444550328785
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,32,12288,0.018248000078731112
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,32,8192,0.00572622236278322
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,32,7168,0.006221333311663733
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,32,10240,0.016883555385801528
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,32,6144,0.006603555546866522
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,64,2560,0.006123555617200003
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,32,5120,0.0058906666106647914
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,32,6144,0.014781332678265043
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,32,4096,0.005552000055710475
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,32,5120,0.014480888843536377
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,32,3584,0.005951111101441913
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,32,4096,0.013423110875818463
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,32,3072,0.005616000129116907
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,32,3584,0.013796444568369122
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,32,2560,0.006234666539563074
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,32,3072,0.013092444174819522
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,64,64,0.0024702221983008915
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,32,2048,0.005879999862776862
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,32,1536,0.00452533322903845
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,32,2560,0.01312444441848331
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,32,1024,0.003887999802827835
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,32,2048,0.011051555474599203
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,64,32,0.00997688869635264
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,32,768,0.0035128887328836652
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,32,512,0.002891555635465516
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,32,1024,0.011699555648697747
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,32,256,0.00282577777074443
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,32,768,0.012110222544935016
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,32,128,0.0028977776981062363
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,32,512,0.01070577816830741
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,32,64,0.002522666628162066
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,32,256,0.01128533316983117
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,32,8192,0.015967110792795818
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,32,32,0.002696000039577484
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,32,7168,0.015496888094478183
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,32,64,0.009023110899660323
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,32,32,0.009993777506881291
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,float16,1,64,768,0.003527111063400904
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,32,1536,0.0123848890264829
VLLM,0.14.1,NVIDIA B200,gemm,vllm_default,fp8,1,32,128,0.010101333260536194
